{"cells":[{"metadata":{},"cell_type":"markdown","source":"# <u>Ensemble Learning</u>\n\nIf we aggregate the predictions of a group of predictors (classifiers / regressors), we will <u> often</u> get better predictions than with the best individual predictor. A group of predictor is called an ensemble, the technique of aggregation is called Ensemble Learning, and an Ensemble Learning algorithm is called an Ensemble method.\n\n<b>Ensemble methods work best when the predictors are as independent from one another as possible. Training the dataset using very different classifiers increases the chance of them making very different types of errors, improving ensemble’s accuracy.</b>\n\n<u>Ensemble methods can be divided into two groups: </u>\n\n- <b> Sequential ensemble methods </b> where the base learners are generated sequentially (e.g. AdaBoost).\n  - The basic motivation of sequential methods is to <u> exploit the dependence between the base learners </u>. The overall performance can be boosted by weighing previously mislabeled examples with higher weight.\n\n- <b> Parallel ensemble methods </b> where the base learners are generated in parallel (e.g. Random Forest).\n  - The basic motivation of parallel methods is to <u> exploit independence between the base learners </u> since the error can be reduced dramatically by averaging."},{"metadata":{},"cell_type":"markdown","source":"## <u>Voting Classifiers</u> :\nTrain a few classifiers (Logistic Regression classifier, SVM classifier, Random Forest classifier, K-Nearest Neighbors classifier etc.) on the training set.\n\n![Screen%20Shot%202019-11-19%20at%2010.47.24.png](attachment:Screen%20Shot%202019-11-19%20at%2010.47.24.png)","attachments":{"Screen%20Shot%202019-11-19%20at%2010.47.24.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":" Aggregate the predictions of each classifier and predict the class that gets the most votes. This majority vote classifier is called a hard voting classifier.\n \n![Screen%20Shot%202019-11-15%20at%2018.25.08.png](attachment:Screen%20Shot%202019-11-15%20at%2018.25.08.png)\n\nThis voting classifier often achieves a higher accuracy than the best classifier in the ensemble. Even if each classifier is a weak learner (it does only slightly better than random guessing), the ensemble can still be a strong learner (achieving high accuracy), provided there are a sufficient number of weak learners and they are sufficiently diverse.","attachments":{"Screen%20Shot%202019-11-15%20at%2018.25.08.png":{"image/png":"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"}}},{"metadata":{"trusted":true},"cell_type":"code","source":"# Load the MNIST dataset\nimport numpy as np\n\ntry:\n    from sklearn.datasets import fetch_openml\n    mnist = fetch_openml('mnist_784', version=1)\n    mnist.target = mnist.target.astype(np.int64)\nexcept ImportError:\n    from sklearn.datasets import fetch_mldata\n    mnist = fetch_mldata('MNIST original')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Split 50,000 instances for training, 10,000 for validation, and 10,000 for testing.\n\nfrom sklearn.model_selection import train_test_split\n\nX_train_val, X_test, y_train_val, y_test = train_test_split( mnist.data, mnist.target, test_size=10000, random_state=42)\nX_train, X_val, y_train, y_val = train_test_split( X_train_val, y_train_val, test_size=10000, random_state=42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Train Random Forest classifier, Extra-Trees classifier, SVM and MLP\n\nfrom sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier\nfrom sklearn.svm import LinearSVC\nfrom sklearn.neural_network import MLPClassifier\n\nrandom_forest_clf = RandomForestClassifier(n_estimators=10, random_state=42)\nextra_trees_clf = ExtraTreesClassifier(n_estimators=10, random_state=42)\nsvm_clf = LinearSVC(random_state=42)\nmlp_clf = MLPClassifier(random_state=42)\n\nestimators = [random_forest_clf, extra_trees_clf, svm_clf, mlp_clf]\nfor estimator in estimators:\n    estimator.fit(X_train, y_train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# .score() method directly calls sklearn.metrics.accuracy_score method.\n\n[estimator.score(X_val, y_val) for estimator in estimators]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Combine the classifiers into an ensemble that outperforms them all on the validation set, using a soft or hard voting classifier.\n\nfrom sklearn.ensemble import VotingClassifier\n\nnamed_estimators = [ (\"random_forest_clf\", random_forest_clf), (\"extra_trees_clf\", extra_trees_clf), (\"svm_clf\", svm_clf), (\"mlp_clf\", mlp_clf)]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"voting_clf = VotingClassifier(named_estimators)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"voting_clf.fit(X_train, y_train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"voting_clf.score(X_val, y_val)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"[estimator.score(X_val, y_val) for estimator in voting_clf.estimators_]","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Linear SVM is far outperformed by the other classifiers. Remove the SVM to see if performance improves."},{"metadata":{"trusted":true},"cell_type":"code","source":"# remove an estimator by setting it to None using set_params()\n\nvoting_clf.set_params(svm_clf=None)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Updated list of estimators\n\nvoting_clf.estimators","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Updated list of trained estimators\n\nvoting_clf.estimators_","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# We can either fit the VotingClassifier again, or just remove the SVM from the list of trained estimators:\n\ndel voting_clf.estimators_[2]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Recheck\n\nvoting_clf.estimators_","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Evaluate the VotingClassifier again:\n\nvoting_clf.score(X_val, y_val)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"A bit better! SVM was hurting performance. Let's try using a soft voting classifier.\n\nIf all classifiers in the ensemble can estimate class probabilities (predict_proba()), we can set \"voting = 'soft'\" to predict the class with the highest class probability, averaged over all the individual classifiers. This is called soft voting. It often achieves higher performance than hard voting because it gives more weight to highly confident votes."},{"metadata":{"trusted":true},"cell_type":"code","source":"# Set voting to \"soft\"\n\nvoting_clf.voting = \"soft\"\nvoting_clf.score(X_val, y_val)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Test set\n\nvoting_clf.score(X_test, y_test)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"[estimator.score(X_test, y_test) for estimator in voting_clf.estimators_]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"markdown","source":"## <u>Bagging and Pasting</u>\nIn Bagging/ Pasting,  same training algorithm is used for every predictor(Classifier/ Regressor), but we train them on different random subsets of the training set. When sampling is performed with replacement, the method is called <b> bagging / bootstrap aggregating </b>. When sampling is performed without replacement, it is called <b> pasting </b>.\n\n![Screen%20Shot%202019-11-15%20at%2018.40.20.png](attachment:Screen%20Shot%202019-11-15%20at%2018.40.20.png)\n\nOnce all predictors are trained, the ensemble can make a prediction for a new instance by aggregating the predictions of all predictors. The aggregation function is generally the statistical mode (most frequent prediction) for classification, or the statistical mean for regression.\n\nEach individual predictor has a higher bias(underfit) because only a subset of the trining set is trained on it, aggregation reduces both bias and variance. The ensemble has a similar bias but a lower variance than a single predictor trained on the original training set.\n\nBagging and Pasting scale very well on different CPU cores and servers, all predictors can be trained in parallel, predictions can also be made in parallel.\n\n### <u>Out-of-Bag Evaluation </u>\n\nWith bagging, at each iteration, the remaining training instances that are not sampled are called out-of-bag (oob) instances. These instances are not same for all predictors.\nSince a predictor never sees the oob instances during training, it can be evaluated on these instances, without the need for a separate validation set or cross-validation. We can evaluate the ensemble itself by averaging out the oob evaluations of each predictor.","attachments":{"Screen%20Shot%202019-11-15%20at%2018.40.20.png":{"image/png":"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"}}},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.tree import DecisionTreeClassifier # Base Classifier for Bagging Method\nfrom sklearn.ensemble import BaggingClassifier\nfrom sklearn.metrics import accuracy_score\n\n# Train an ensemble of 500 Decision Tree classifiers.\n# BaggingClassifier automatically performs soft voting if the base classifier has a predict_proba() method.\n# To train each predictor on a random subset of the input features, use max_features(0-1). Useful if training set has high dimentional features.\n# \"n_estimators = 500\" was taking lot of CPU time, I used \"n_estimators = 10\" for demonstration purpose.\n# bootstrap = True (Bagging)\nclf_bagging = BaggingClassifier( DecisionTreeClassifier(), n_estimators = 10, max_samples = 0.8, bootstrap = True, oob_score=True, n_jobs = -1)\n# bootstrap = False (Pasting)\nclf_pasting = BaggingClassifier( DecisionTreeClassifier(), n_estimators = 10, max_samples = 0.8, bootstrap = False, n_jobs = -1)\n\nclf_bagging.fit(X_train, y_train)\nclf_pasting.fit(X_train, y_train)\n\ny_pred_bagging = clf_bagging.predict(X_val)\ny_pred_pasting = clf_bagging.predict(X_val)\n\n\nclf_bagging.oob_score_, accuracy_score(y_val, y_pred_bagging), accuracy_score(y_val, y_pred_pasting)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# The decision function returns the class probabilities (if base estimator has a predict_proba() method) for each training instance. \nclf_bagging.oob_decision_function_[112]\n#y_train[112]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# import matplotlib as mpl\n\n# some_digit = X_train[112]\n# some_digit_image = some_digit.reshape(28, 28)\n# plt.imshow(some_digit_image, cmap = mpl.cm.binary, interpolation=\"nearest\")\n# plt.axis(\"off\")\n\n# plt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## <u> Random Forests </u>\nRandom Forest is an ensemble of Decision Trees.\n\nInstead of building a BaggingClassifier and passing it a DecisionTreeClassifier, we can instead use the RandomForestClassifier class, which is more convenient and optimized for Decision Trees.\n\nRandomForestClassifier has all the hyperparameters of a DecisionTreeClassifier (to control how trees are grown), plus all the hyperparameters of a BaggingClassifier to control the ensemble itself."},{"metadata":{},"cell_type":"markdown","source":"In an <b> extremely randomized trees </b> algorithm randomness goes one step further: the splitting thresholds are randomized. Instead of looking for the most discriminative threshold, thresholds are drawn at random for each candidate feature and the best of these randomly-generated thresholds is picked as the splitting rule. This usually allows reduction of the variance of the model a bit more, at the expense of a slightly greater increase in bias."},{"metadata":{"trusted":true},"cell_type":"code","source":"#  Trains a Random Forest classifier with 500 trees (each limited to maximum 16 nodes), using all available CPU cores:\n\nfrom sklearn.ensemble import RandomForestClassifier\n\nclf_randomforest = RandomForestClassifier(n_estimators=10, n_jobs=-1)\nclf_randomforest.fit(X_train, y_train)\ny_pred_rf = clf_randomforest.predict(X_val)\n\naccuracy_score(y_val, y_pred_rf)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### <u>Feature Importance</u>\n\nIf we look at a single Decision Tree, important features are likely to appear closer to the root of the tree, while unimportant features will often appear closer to the leaves (or not at all). It is possible to get an estimate of a feature’s importance by computing the average depth at which it appears across all trees in the forest."},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.datasets import load_iris\niris = load_iris()\nrnd_clf = RandomForestClassifier(n_estimators=500, n_jobs=-1)\nrnd_clf.fit(iris[\"data\"], iris[\"target\"])\nfor name, score in zip(iris[\"feature_names\"], rnd_clf.feature_importances_):\n    print(name, score)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## <u>Boosting</u>\nBoosting refers to a family of algorithms that are able to convert weak learners to strong learners. The main principle of boosting is to fit a sequence of weak learners− models that are only slightly better than random guessing, such as small decision trees to weighted versions of the data. More weight is given to examples that were misclassified by earlier rounds.\n\nThe predictions are then combined through a weighted majority vote (classification) or a weighted sum (regression) to produce the final prediction. \n\nBoosting technique cannot be parallelized (or only partially) because each predictor can only be trained after the previous predictor has been trained and evaluated. As a result, it does not scale as well as bagging / pasting."},{"metadata":{},"cell_type":"markdown","source":"### <u> AdaBoost </u>\n\nThe predictors(classifier/ regressor) fit the training set in sequence. The next predictor corrects its predecessor by paying more attention to the training instances that the predecessor underfitted. This results in new predictors focusing more and more on the hard cases.\n\nTo build an AdaBoost classifier, each instance's weight is set to an initial value. A base classifier (eg. Decision Tree) is trained and makes predictions on the training set. The relative weight of misclassified training instances is then increased. The second classifier is trained on the training set using the updated weights and again it makes predictions on the training set and update the weights. The algorithm stops when the desired number of predictors is reached, or when a perfect predictor is found.\n\n![Screen%20Shot%202019-11-17%20at%2023.41.39.png](attachment:Screen%20Shot%202019-11-17%20at%2023.41.39.png)\n\nThis sequential learning technique is similar to Gradient Descent, except that instead of tweaking a single predictor’s parameters to minimize a cost function, AdaBoost adds more predictors to the ensemble, gradually making it better.\n\n<b><u> Algorithm </u></b>:\n\n- Each instance weight $w^{(i)}$ is initially set to $\\frac{1}{m}$ (m = number of instances) . The first predictor is trained and its weighted error rate $r_1$ is computed on the training set:\n\n$$\\text{ Weighted error rate for $j^{th}$ predictor:   } r_j = \\frac{\\hat{y_j}^i \\neq y^i}{\\sum_{i=1}^{m} w^i} $$\n\nwhere ${y_j}^i$ is the $j^{th}$ predictor’s prediction for the $i^{th}$ instance.\n\n- The predictor’s weight $\\alpha_j$ is then computed using its weighted error rate , where $\\eta$ is the learning rate hyperparameter (defaults to 1).\n  - Predictor guessing accurately - Higher weight\n  - Predictor guessing randomly - Weight close to 0\n  - Predictor guessing mostly wrong - Negative weights\n\n$$\\text{ Predictor weight:   } \\alpha_j = \\eta \\log\\frac{1 - r_j}{r_j} $$\n\n- The instance weights are updated using above equation and the misclassified instance' weights are boosted.\n\n$$\\text{ Weight update rule:   } w^{i}  \\leftarrow \n  \\begin{cases}\n    w^i       & \\quad \\text{if }  \\hat{y_j}^i = y^i\\\\\n    w^i \\exp(\\alpha_j)  & \\quad \\text{if } \\hat{y_j}^i \\neq y^i\n  \\end{cases}\n$$\n\n- All the instance weights are normalized (divided by $\\sum_{i=1}^{m} w^i$).\n\nA new predictor is trained using the updated weights, and the whole process is repeated (the new predictor’s weight is computed, the instance weights are updated, then another predictor is trained. \n\nTo make predictions, AdaBoost simply computes the predictions of all the predictors and weighs them using their predictor weights $\\alpha_j$. The predicted class is the one that receives the majority of weighted votes.","attachments":{"Screen%20Shot%202019-11-17%20at%2023.41.39.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"Scikit-Learn uses a multiclass version of AdaBoost - SAMME. When there are just two classes, SAMME is equivalent to AdaBoost. If the predictors can estimate class probabilities (if they have a predict_proba() method), Scikit-Learn can use a variant of SAMME called SAMME.R, which relies on class probabilities rather than predictions and generally performs better.\n\nThe following code using Scikit-Learn’s AdaBoostClassifier class (as you might expect, there is also an Ada BoostRegressor class). A Decision Stump is a Decision Tree with max_depth=1—in other words, a tree composed of a single decision node plus two leaf nodes. This is the default base estimator for the AdaBoostClassifier class:"},{"metadata":{"trusted":true},"cell_type":"code","source":"# Train a classifier based on 200 Decision Trees \n\nfrom sklearn.ensemble import AdaBoostClassifier\nada_clf = AdaBoostClassifier( DecisionTreeClassifier(max_depth=1), n_estimators=200, algorithm=\"SAMME.R\", learning_rate=0.5)\nada_clf.fit(X_train, y_train)\ny_pred_ada = ada_clf.predict(X_val)\naccuracy_score(y_val, y_pred_ada)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"If AdaBoost ensemble underfits the training data, you can try increasing the number of estimators or reducing the regularization hyperparameters of the base estimator. You may also try slightly increasing the learning rate."},{"metadata":{},"cell_type":"markdown","source":"### Gradient Boosting\nGradient Boosting works by sequentially adding predictors to an ensemble, each one correcting its predecessor. However, instead of tweaking the instance weights at every iteration like AdaBoost, Gradient Boosting tries to fit the new predictor to the residual errors made by the previous predictor."},{"metadata":{"trusted":true},"cell_type":"code","source":"import numpy as np\n\nnp.random.seed(42)\nX = np.random.rand(100, 1) - 0.5\ny = 3*X[:, 0]**2 + 0.05 * np.random.randn(100)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.tree import DecisionTreeRegressor\ntree_reg1 = DecisionTreeRegressor(max_depth=2)\ntree_reg1.fit(X, y)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Now train a second DecisionTreeRegressor on the residual errors made by the first\npredictor:"},{"metadata":{"trusted":true},"cell_type":"code","source":"y2 = y - tree_reg1.predict(X)\ntree_reg2 = DecisionTreeRegressor(max_depth=2)\ntree_reg2.fit(X, y2)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Then we train a third regressor on the residual errors made by the second predictor:"},{"metadata":{"trusted":true},"cell_type":"code","source":"y3 = y2 - tree_reg2.predict(X)\ntree_reg3 = DecisionTreeRegressor(max_depth=2)\ntree_reg3.fit(X, y3)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Now we have an ensemble containing three trees. It can make predictions on a new instance simply by adding up the predictions of all the trees:"},{"metadata":{"trusted":true},"cell_type":"code","source":"\nX_new = np.array([[0.8]])\n\ny_pred = sum(tree.predict(X_new) for tree in (tree_reg1, tree_reg2, tree_reg3))\ny_pred","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"The following code creates the same ensemble as the previous one:"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.ensemble import GradientBoostingRegressor\ngbrt = GradientBoostingRegressor(max_depth=2, n_estimators=3, learning_rate=1.0)\ngbrt.fit(X, y)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"The learning_rate hyperparameter scales the contribution of each tree. If we set it to a low value(0.1) we will need more trees in the ensemble to fit the training set, but the predictions will usually generalize better. This is a regularization technique called shrinkage.\n\nIf Gradient Boosting ensemble overfits the training set, we should try decreasing the learning rate. We could also use early stopping to find the right number of predictors."},{"metadata":{"trusted":true},"cell_type":"code","source":"gbrt_slow = GradientBoostingRegressor(max_depth=2, n_estimators=200, learning_rate=0.1, random_state=42)\ngbrt_slow.fit(X, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import matplotlib.pyplot as plt\n\n\ndef plot_predictions(regressors, X, y, axes, label=None, style=\"r-\", data_style=\"b.\", data_label=None):\n    x1 = np.linspace(axes[0], axes[1], 500)\n    y_pred = sum(regressor.predict(x1.reshape(-1, 1)) for regressor in regressors)\n    plt.plot(X[:, 0], y, data_style, label=data_label)\n    plt.plot(x1, y_pred, style, linewidth=2, label=label)\n    if label or data_label:\n        plt.legend(loc=\"upper center\", fontsize=16)\n    plt.axis(axes)\n\nplt.figure(figsize=(11,11))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\nplt.figure(figsize=(11,4))\n\nplt.subplot(121)\nplot_predictions([gbrt], X, y, axes=[-0.5, 0.5, -0.1, 0.8], label=\"Ensemble predictions\")\nplt.title(\"learning_rate={}, n_estimators={}\".format(gbrt.learning_rate, gbrt.n_estimators), fontsize=14)\n\nplt.subplot(122)\nplot_predictions([gbrt_slow], X, y, axes=[-0.5, 0.5, -0.1, 0.8])\nplt.title(\"learning_rate={}, n_estimators={}\".format(gbrt_slow.learning_rate, gbrt_slow.n_estimators), fontsize=14)\n\n#save_fig(\"gbrt_learning_rate_plot\")\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"In order to find the optimal number of trees, you can use early stopping (see Chap‐ ter 4). A simple way to implement this is to use the staged_predict() method: it returns an iterator over the predictions made by the ensemble at each stage of train‐ ing (with one tree, two trees, etc.). The following code trains a GBRT ensemble with 120 trees, then measures the validation error at each stage of training to find the opti‐ mal number of trees, and finally trains another GBRT ensemble using the optimal number of trees:"},{"metadata":{"trusted":true},"cell_type":"code","source":"import numpy as np\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import mean_squared_error\nX_train, X_val, y_train, y_val = train_test_split(X, y)\ngbrt = GradientBoostingRegressor(max_depth=2, n_estimators=120)\ngbrt.fit(X_train, y_train)\nerrors = [mean_squared_error(y_val, y_pred) for y_pred in gbrt.staged_predict(X_val)]\nbst_n_estimators = np.argmin(errors)\ngbrt_best = GradientBoostingRegressor(max_depth=2,n_estimators=bst_n_estimators)\ngbrt_best.fit(X_train, y_train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"min_error = np.min(errors)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.figure(figsize=(11, 4))\n\nplt.subplot(121)\nplt.plot(errors, \"b.-\")\nplt.plot([bst_n_estimators, bst_n_estimators], [0, min_error], \"k--\")\nplt.plot([0, 120], [min_error, min_error], \"k--\")\nplt.plot(bst_n_estimators, min_error, \"ko\")\nplt.text(bst_n_estimators, min_error*1.2, \"Minimum\", ha=\"center\", fontsize=14)\nplt.axis([0, 120, 0, 0.01])\nplt.xlabel(\"Number of trees\")\nplt.title(\"Validation error\", fontsize=14)\n\nplt.subplot(122)\nplot_predictions([gbrt_best], X, y, axes=[-0.5, 0.5, -0.1, 0.8])\nplt.title(\"Best model (%d trees)\" % bst_n_estimators, fontsize=14)\n\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"gbrt = GradientBoostingRegressor(max_depth=2, warm_start=True, random_state=42)\n\nmin_val_error = float(\"inf\")\nerror_going_up = 0\nfor n_estimators in range(1, 120):\n    gbrt.n_estimators = n_estimators\n    gbrt.fit(X_train, y_train)\n    y_pred = gbrt.predict(X_val)\n    val_error = mean_squared_error(y_val, y_pred)\n    if val_error < min_val_error:\n        min_val_error = val_error\n        error_going_up = 0\n    else:\n        error_going_up += 1\n        if error_going_up == 5:\n            break  # early stopping","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(gbrt.n_estimators)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(\"Minimum validation MSE:\", min_val_error)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"markdown","source":"### Stacking\n\nStacking is an ensemble learning technique that uses predictions from multiple models (for example decision tree, knn or svm) to build a new model. This model is used for making predictions on the test set.\n\n![Screen%20Shot%202019-11-17%20at%2014.20.06.png](attachment:Screen%20Shot%202019-11-17%20at%2014.20.06.png)\n\nFirst, the training set is split in two subsets. The first subset is used to train the predictors in the first layer. Next, the predictors in the first layer are used to make predictions on the second(hold-out) set. Now (in example above) for each instance in the hold-out set there are four predicted values. A new training set is created using these predicted values as input features and keeping the target values. The blender is trained on this new training set, it learns to predict the target value where inputs are the the first layer’s predictions.\n\nIt is possible to train several different blenders on the top of one another (e.g., one using Linear Regression, another using Random Forest Regression etc). The training set should be divided equal to the number of layers(see image above).","attachments":{"Screen%20Shot%202019-11-17%20at%2014.20.06.png":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAnAAAAGlCAYAAAB6CMuiAAAYKWlDQ1BJQ0MgUHJvZmlsZQAAWIWVWQk4ld233+97ZscxHPM8z/M8ZZ7HzFPEMR+zYyZjUlIqSoaoJCGlCZEUqVSSKNGoJJT6UJJM90X1fff73/vc5+7n2e/5nbXXWu9v7732sM4BgIeTEh0dDjMBEBEZR3M0NxJ09/AUxI0BOsAHmIEkYKP4x0Yb2tvbgP+1fB8C0PrnoNy6r/9d738szAGBsf4AQPYI9guI9Y9A8BUA0Nz+0bQ4ADD9iFwkMS56HX9DMCsNIQgAFr+Ogzcx7zr228SKGzrOjsYINgEAT0+h0IIBYFj3L5jgH4z4YYhG2siRAdRIRDUHwXr+IZQAALi7ER3ZiIiodTyDYEm/f/gJ/m8+/f74pFCC/+DNvmwUvAk1Njqckvz/HI7/u0SEx/9+hzBS6UNoFo7rfUbGrTYsynod0yO4PdLPbiuCyQjuoQZs6K/j5yHxFi6/9Kf9Y42RMQPsAMAggGJijWBkLGH2+DAXw19YmULbsEX0YTtqnKXzL+xHi3L85R9OCIw1dfqNQwItbX753BMZbvcbVwZRzSwRjEQafCUlxNltkyfcnUB1tUMwA4L7Y8OcrH/pv04JMbb7rUOLd1znLIrgb0E0M8dNHRRnROzvfqHk/SkbHDgRbBAX4myxaYtyD4x1t/nNLSDQxHSTAyogMNLlF2cUEl1Gjr9sc6PD7X/poyoDw80dN8cZdSE2wem37UAcEmCb44B6F0qxst/kj/oeHWfvvMkNjQY2wBiYAEEQj1Q/EAVCAbVvumUa+bbZYgYogAaCQSCQ+yX5beG20RKJPJ1ACviMoEAQ+8fOaKM1ECQg8pU/0s2nHAjaaE3YsAgDHxAcgeZG66F10DbI0wCpymhNtNZvO0HG32/FmmJNsBZYM6zUHx7+COtwpNIA9T9lf1tiPmAeY95hnmJGMSPAGmkNRPq8zjDyT89cwfsNL7+++1Czaf9iLghswShiZ/ard36I9dRvHbQ4wloNbYTWRfgj3NHsaG4gh1ZFemKI1kf6poZI/8kw/g+Lv8fy3+9b5/fPPv6SM0gzqP1i4feHv/EfrX97Mf7HGAUgn9b/1kTtQV1G3UV1ou6h2lEtQBB1A3UV1Yu6vo7/RML7jUj4/TbHDW5hiB/qbx3Fs4pTisv/8XbKLwa0jfkGcYFJcesLwjgqOplGDQ6JEzREduRAQctIf3lZQWVFJS0A1vf3ze3jq+PGvg2xP/pbFpoKgIYAIrz1tyxwCIC2V8iWRve3THwXEvJoAO75+sfTEjZl6PUHBjk5GJGVwQX4gQhyfsgBZaAOdIABMAVWYCtwBh5gOzLqISACYZ0IdoAskAvywUFwBJSB4+AUqAXnwCXQAtpBJ7gDHoB+8BS8QGJjHHwCM+A7WIIgCAeRIBaICxKAxCAZSBnShPQgU8gGcoQ8IF8oGIqE4qEd0E4oHyqEyqCTUB10EWqFOqF70GNoBHoLTUFz0E8YBdPDrDAfLA4rwJqwIWwNO8PecDAcA6fAOXABXAJXwQ1wM9wJP4CfwqPwJ3geBVBEFDtKCCWH0kQZo7aiPFFBKBoqHbUXVYyqQp1HtSFzPYgaRU2jFtFYNAtaEC2HxKcF2gXtj45Bp6P3ocvQtehmdDd6EP0WPYNexZAwvBgZjDbGEuOOCcYkYnIxxZgaTBPmNrKixjHfsVgsO1YCq4GsTQ9sKDYVuw9bgW3E3sQ+xo5h53E4HBdOBqeL24qj4OJwubhSXAPuBm4AN477gSfiBfDKeDO8Jz4Sn40vxtfjO/AD+An8EoGJIEbQJmwlBBCSCQcI1YQ2wiPCOGGJjplOgk6XzpkulC6LroTuPN1tupd0X4lEojBRi+hApBIziSXEC8Qe4lviIj2ZXpremN6LPp6+gP4M/U36EfqvJBJJnGRA8iTFkQpIdaRbpNekHwwsDPIMlgwBDBkM5QzNDAMMXxgJjGKMhozbGVMYixkvMz5inGYiMIkzGTNRmNKZyplamZ4xzTOzMCsxb2WOYN7HXM98j3mSjCOLk03JAeQc8inyLfIYC4pFhMWYxZ9lJ0s1y22WcVYsqwSrJWsoaz7rOdY+1hk2MpsqmytbEls523W2UXYUuzi7JXs4+wH2S+xD7D85+DgMOQI58jjOcwxwLHDycBpwBnLu5WzkfMr5k0uQy5QrjOsQVwvXK240tzS3A3cidyX3be5pHlYeHR5/nr08l3ie88K80ryOvKm8p3h7eef5+PnM+aL5Svlu8U3zs/Mb8IfyH+bv4J8SYBHQE6AKHBa4IfBRkE3QUDBcsESwW3BGiFfIQihe6KRQn9CSsISwi3C2cKPwKxE6EU2RIJHDIl0iM6ICoraiO0TPij4XI4hpioWIHRW7K7YgLiHuJr5bvEV8UoJTwlIiReKsxEtJkqS+ZIxkleQTKayUplSYVIVUvzQsrSYdIl0u/UgGllGXocpUyDyWxchqyUbKVsk+k6OXM5RLkDsr91aeXd5GPlu+Rf6LgqiCp8IhhbsKq4pqiuGK1YovlMhKVkrZSm1Kc8rSyv7K5cpPVEgqZioZKldVZlVlVANVK1WH1VjUbNV2q3WprahrqNPUz6tPaYhq+Goc03imyappr7lPs0cLo2WklaHVrrWora4dp31J+y8dOZ0wnXqdyS0SWwK3VG8Z0xXWpeie1B3VE9Tz1TuhN6ovpE/Rr9J/ZyBiEGBQYzBhKGUYathg+MVI0Yhm1GS0YKxtnGZ80wRlYm6y16TPlGzqYlpm+tpM2CzY7KzZjLmaear5TQuMhbXFIYtnlnyW/pZ1ljNWGlZpVt3W9NZO1mXW72ykbWg2bbawrZVtke1LOzG7SLuWrWCr5daira/sJexj7K85YB3sHcodPjgqOe5wvOvE4uTjVO/03dnI+YDzCxdJl3iXLldGVy/XOtcFNxO3QrdRdwX3NPcHHtweVI+rnjhPV88az/ltptuObBv3UvPK9RrylvBO8r63nXt7+PbrPow+FJ/LvhhfN99632XKVkoVZd7P0u+Y34y/sf9R/08BBgGHA6YCdQMLAyeCdIMKgyaDdYOLgqdC9EOKQ6apxtQy6myoRejx0IWwrWFnwtbC3cIbI/ARvhGtkeTIsMjuKP6opKjH0TLRudGjMdoxR2JmaNa0mlgo1jv2ahwrctXpjZeM3xX/NkEvoTzhR6Jr4uUk5qTIpN5k6eS85IkUs5TTqehU/9SuHUI7sna8TTNMO5kOpfuld2WIZORkjGeaZ9Zm0WWFZT3MVswuzP62021nWw5fTmbO2C7zXWdzGXJpuc926+w+vge9h7qnL08lrzRvdW/A3vv5ivnF+cv7/Pfd36+0v2T/WkFQQd8B9QOVB7EHIw8OHdI/VFvIXJhSOFZkW9R8WPDw3sPfjvgcuVesWnz8KN3R+KOjJTYlV0tFSw+WLpeFlD0tNypvPMZ7LO/YQkVAxUClQeX543zH84//PEE9MXzS/GRzlXhV8SnsqYRTH6pdq++e1jxdV8Ndk1+zcibyzGitY213nUZdXT1v/YGz8Nn4s1MNXg3950zOXT0vd/5kI3tj/gVwIf7Cx4u+F4cuWV/quqx5+fwVsSvHmlia9jZDzcnNMy0hLaNXPa4+brVq7WrTaWu6Jn/tTLtQe/l1tusHOug6cjrWbqTcmL8ZfXO6M7hzrMun68Ut91tPuh26+25b3+65Y3bn1l3Duzd6dHva72nfa72veb/lgfqD5l613qaHag+b+tT7mh9pPLrar9Xf9njL444B/YHOQZPBO08snzx4avf08ZDL0PAzr2ejwwHDkyPhI7PPE54vvch8iXm59xXTq+LXvK+r3ki9aRxVH73+1uRt7zundy/G/Mc+vY99vzye84H0oXhCYKJuUnmyfcpsqv/jto/jn6I/LU3nfmb+fOyL5Jcrfxn81TvjPjM+S5tdm9v3levrmW+q37rm7edff4/4vrSw9wfXj9pFzcW7P91+TiwlLuOWS1akVtpWrVdfrkWsrUVTaJSNqwAKqXBQEABzZwAgeQDAguRxdAyb+devgoLW0451XRJyi9mC3LaKQB9EhtyhWhiGI+AxVCBqDp2PUcSMYitwoXgTgjgdAxGmR5GYGWQYLZlozCfJr1j52fzYL3GiuXy5b/IK8OXxzwp6Cz0Q0RY9Lc4qkSk5IW0n0yjHIO+vcFlxSVlHJVb1uFq3+luNRS16bW4d6S2auiZ6dvqeBiGGCUa5xsUmtaZtZvfNn1tMWi5Yo22YbHntJLYq2Ws7GDlaOtk5O7q4uLq5ubt7eHh6em7z9PL09tzu7uPq60ix9TPz1wtQC5QOEghmCcGFLFG/hL4NexJ+F1mVZ6MqovfHJNMosYZxXHFf4jsTjiZGJVkliySvpDxLbdyxJ803XSODAVlb17IKs0N26uaw5Ezu6sgt2h2yZ0see95KPnqf3v5zBzQPXjq0UiRwWOaIfLHiUaUSlVLVMrVytWPqFdqVZscDT5ScHD7FVm142rsm8kxKbW7dofrys6cbGs+1nr/VOHDh8yWhy9FX+pulWsKvlrQ2tz26NtG+2sF+Q+mma2dh12S3xe3yOw/vvu2ZuY99INZr/jCgL/ZReL/LY40B/kG6wcUnY08fDt141jbcPnLjeeeLjpeNrw69Dn9jNMo1Ove2/13rWO378vGDH3ZNJE9GTPl+tP2kMk2e/vT5zpfqv3JnQmft5lS/Cn+Tmvf+3vFDcfHwzzfLXCvuq9Vra+txAoiAB7klOiK5TgP4AElAUdBNmAfOhudQ0agf6D0YIcxtbBxOHvcV30WooEsjBtC7k5wY3Bn9mOKZ88m1LP2sP9glOLw5i7ge8ZB4bfj28fcJkoQchA+J9IsRxU0lEiRrpB5Lf5NlkpOUV1XQUtRSUlGWUuFXZVKD1L6pjyOnVY9Wq3adTtmWfN1UvVD9bQZ2hkZGGsbyJqKm3GZM5ljzJYsZy3GrYetemw7bC3ZVW4vscxxiHSlO9s56LjKunG4Yt1n3lx49npe3HffK847d7u1j6itLYaH88Hvl3xlQHbgnKCzYNkSBykz9Gvo0rDm8JCI50j1KPZocPRVzg1YUGxCnFo+JH0o4nRifZJrMmjyWcik1c4ddGm/ax/S2jP2ZoVmO2SZIZGjvUs9V3C2zRyxPYC9XPnkfcT96/0rB9wOzB+cOLRbhDnMekSzWOGpSYl+6rSy4nHYsrWJ3ZeHxYyfOnLxaNXBq8bRUjdeZ/Nqmuuf1qw1C50zPhzTuv9By8ctltSu7mh63kK7qtlLbSq89aF/rULsRfrO68+Ut5m6D29Q7+Xfre3ruTT0g9ao89OzLftTQ/2wAO6j6xOdpzlD1s+7hD8/pXii8dH6V/Lryzd3RhXdKY7T3l8fnJmQng6dOfXwzzfPZ/cuxv2ZmE77KzZMX6Bbhn5+Wr61Sf80/HeAAssACyXiOgvsQFjKHDkFjsC58EkVC7ULj0IUYccxNbACOjLuH30OwoxOgWyQ+ob9KOs1QyljIdIC5kFzGcpq1ma2H/TXHIheZW47HjJfCt4P/qMB5wS6hJ8LjIp9F58RmkFvTsGSX1GnpnTKesgpykNyAfLVCoqKVkqDSgnKfSo1qupqLuqwGrDGseU4rW9tVR1pnZUu/7im9RH0rAwGDecNeo9PG6SaupvJmGLOX5lcs9lr6WqlbE61HbZps8+y8kZ0CYz/i0OCY6eTkLOb83aXHtcwtzF3Hg+jxwvPsthQvC2927/fbL/pk+FpTOChjfuf9UwLMApkDXwTVBMeE6FDR1L7Qo2G+4VLhsxFXIzOjzKIJ0b0x+2hWsfjY23HZ8XrxSwktibFJ8klTydUpPqncqU92FKRZpMPpHRnpmRZZfFlL2aM7e3Iu7irPzdkdscc1T2+veD4pf37fy/23CuoPHD6YdSixkFYUfRi5FhTHHI0piS6NLKOW+x5zqrCqtDnufSL5ZEXV7VNfTrPVaJyxqXWsc6jfdja14cq5pUbzC0UX31yWuZLQ1NlCvOrUWtL2ol3oenjH9ZssnaFdt7p5bsfd6esRv5d2/0mv9MPsvrF+18dDg/5P5of2DHOPnHth8HLodeao7Tun9wc/LEwdnr4967wwsj7/m7/DrResOgCnzQBwPQyAE5Kpni4AQKwWOT+2AGBPAsBZC8BcpQC6Hg0gL8k/5wc/METOjp2gGtxGdg8ssn9YQmHQfqgRyfW+wRywDuwD74Rr4T74K4obZYgKQR1EMvB3aCJaHU1B70e3oicwbBhTTDySdQ1j6bGG2ETseewkThjng6vEvcYL40PwF/ArBBvCCcJ3Onu6c0QSMZI4QK9Jf5JEJCWQxhgcGDoZlRmrmbiYDjITmHeRYXI2C4Ylj5WJtZRNjO0yuwn7MEcUJ56zmsuY6z33bh45nqe8GXyyfC/5CwRMBFYE24RShPVEMCKPRI+JhYnrSpAlPkp2S1VJZ8sEytrIacnLKygo6im5KIer7ES2/Cb1QY3vWnza5joJW+p03+hzGbgZlhq9MZE0jTe7Y8FtGWx1xPqoTYKtge2aXefWffahDlTHHKcLzu9dud2c3As8ereRvBy8i7cP+zJSVPzM/V0CAgIzgs4GT1KVQrPCBiMkkch7HqNBK479Ee+W0JD4OZkjRTHVaIdHWkZ6ayYhKyT7YY76rqrdjHvS8ybyDffl7G8qGD3IcMiu8MJh1SO3j9qVPCyzKL9T4VD540RPVUf1pZqjtSn11IZt5w0vsF18e/lcU0bL9lbPazuut9xY7NLqjrizt6f0fnVvY19H/+OBiaf4Z7oj+198e+052jRGHKdMtH3ET0t8AX9VzPLPlXzjnW9eiFhU+7m83Lzqs7F/iAJrEAOKQTt4B+EhecgZSoGqkEx/FuaCjeAw+Ah8E/6E5OzGyGlSgepFLaFl0F7oAnQneh4jjaFgSjCPsUSsOXYXthuHxVnhDuCG8aL4WPxtAi8hkTBEp0V3gkhHTCRO0LvTPyQZk9oZNBmaGdUZW5n0me4gOeoIOZA8x5LNyspay6bPNsIez8HK0czpyQVzNXB78BB42nljkbme5D8jQBWUF/wu1Cl8QMRHVEWMTuy9eJdEtWSeVKy0r4yDrKncFnkNBTVFdSVtZSMVW9VtapHquRo1mo+0VnVUt0TpntebM9AyzDEaNJEwzTB7YaFjWWG1YmNvW2R3f+uyg7xjgFOl83Nkjr3cT3p83KbmtdN70EfMN5bS7rcaoBuYFtQZQqC6hp4OW4iwiTwVtRzjQbsaxxW/I+F5kmJyasr11J9p2ulZGX1ZwtnJOwd3KeYW7P6SZ7e3Pn9pv0HBjgNNB+cLTYqqjhCKaUeHS/XKTh3DV0RVDp3QPVlzirU6rwZ7pqCOv/5Kg+25scaki8RLx6+oNt1v8b0637annfd60w23TrirqZt6h/du373MB6q9H/tO9W8bYB688dT/GRguf6714tWr3W+UR9+82/deZ3x6onLK7uP89J7Pi39ZzuyavTjX93Xy29p3zgWVH86LO37WL31c0Vw9sjH/UsAZZIA6MAhWISlk9jOhBmgYxsKqsB98CO5EbhEiKFdUHuo66itaCu2LLkUPYhgxNph8zAMsCeuILcW+w8nh0nCP8OL4LPxbgjnhEp0oXTmRnXiEnp2+jCRAqmFQZGhntGF8g9w3GJkbyLbkWZZSVhPWObZT7K4cRI5OzhQuda7v3K08GbyWfBzIXF8XOCJIQ24gqiLcomjk7BkTH5EYkHyEZOZPZV7LfpJbViAryipZISu6SLVD7bOGoKabVqH2wBY2XW+9ev0lQ3ujehOCaYTZMwsryzvWNjbDdlR74FDutMX5nWuBu77H/LaL3jQfdd85v/IAmcBzwdIhNaHiYXURCpGt0eYxw7ER8diEqiTD5DepSWnY9IJM1qzSncI553J1dj/M88+H9p0t8DqIPVRexH/4SDHuaGLJRJln+WCFe+W3E3VVgdW403trvte61zWdZW2IOzfUqHWh8hLmcuSVkWbzltZWxbb6dtHr5TcYbqZ1frzl1t19R/nuyXvk+zkPFh6G973v9348Muj25NmQ87N7I8rPC198eqX3uuDNq7dy79LH+sdFPiRNPJwS+Zjw6eb06helv6xmPGY95uy+bvkmMo+bf/e9bSHzh+6PmcWsn+SfJ5YISzFLI8tGy6XLkysaK7tWnqyKrFJXz63OrKmuJa1dW5//2CAV5Y3jA6I3AgDzem3tqzgAuEIAVg6trS1Vra2tnEKSjJcA3Azf/G9n46xhAuDYrXV0J2Xyx7//Y/kvFR3FLBBrhn0AAAGdaVRYdFhNTDpjb20uYWRvYmUueG1wAAAAAAA8eDp4bXBtZXRhIHhtbG5zOng9ImFkb2JlOm5zOm1ldGEvIiB4OnhtcHRrPSJYTVAgQ29yZSA1LjQuMCI+CiAgIDxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+CiAgICAgIDxyZGY6RGVzY3JpcHRpb24gcmRmOmFib3V0PSIiCiAgICAgICAgICAgIHhtbG5zOmV4aWY9Imh0dHA6Ly9ucy5hZG9iZS5jb20vZXhpZi8xLjAvIj4KICAgICAgICAgPGV4aWY6UGl4ZWxYRGltZW5zaW9uPjYyNDwvZXhpZjpQaXhlbFhEaW1lbnNpb24+CiAgICAgICAgIDxleGlmOlBpeGVsWURpbWVuc2lvbj40MjE8L2V4aWY6UGl4ZWxZRGltZW5zaW9uPgogICAgICA8L3JkZjpEZXNjcmlwdGlvbj4KICAgPC9yZGY6UkRGPgo8L3g6eG1wbWV0YT4Kr5fbgQAAQABJREFUeAHsnQWYJMX5xuvQ4O5yeHAJ7kdwdwtwHO5uwQ8JcEBwwhEOd4fDHYI7h7s7BCc4/f9+X/416entmZ2Znd2xt55nd2Zaq9+qrnrr036JlaAiBISAEBACQkAICAEh0DIIjNQyNVVFhYAQEAJCQAgIASEgBBwBETh1BCEgBISAEBACQkAItBgCInAt1mCqrhAQAkJACAgBISAERODUB4SAEBACQkAICAEh0GIIiMC1WIOpukJACAgBISAEhIAQEIFTHxACQkAICAEhIASEQIshIALXYg2m6goBISAEhIAQEAJCQAROfUAICAEhIASEgBAQAi2GgAhcizWYqisEhIAQEAJCQAgIARE49QEhIASEgBAQAkJACLQYAiJwLdZgqq4QEAJCQAgIASEgBETg1AeEgBAQAkJACAgBIdBiCIjAtViDqbpCQAgIASEgBISAEBCBUx8QAkJACAgBISAEhECLISAC12INpuoKASEgBISAEBACQkAETn1ACAgBISAEhIAQEAIthoAIXIs1mKorBISAEBACQkAICAEROPUBISAEhIAQEAJCQAi0GAIicC3WYKquEBACQkAICAEhIARE4NQHhIAQEAJCQAgIASHQYgiIwLVYg6m6nYXAa6+9Fp5++unch/7hhx/CHXfcEV588cXc/dooBISAEBAC7YuACFz7tq2erA0QOOKII8Jmm23W5UnefPPNMMsss4QVVlghzDHHHGHPPffscow2CAEhIASEQPsiIALXvm2rJ2tjBCB1k0wySYDIDRs2LJx44onh5ptvbuMn1qMJASEgBIRAGgERuDQa+i4EWgCBESNGhIceeij89a9/DdNPP33YaqutwmyzzRaGDh3aArVXFYWAEBACQqAeCIjA1QNFXUMI9CECTz75pN9trrnmKtwVNerjjz9e+K0vQkAICAEh0N4IiMC1d/vq6doQgffffz+MPvrobgMXH2/22WcPn376afj111/jJn0KASEgBIRAGyMgAtfGjatHa08ERh111PDLL7+En3/+ufCAP/74Yxh55JH9r7BRX4SAEBACQqBtERCBa9um1YO1KwJTTTVV+P3338PLL79ceMTnn38+sL1fv36FbfoiBISAEBAC7YuACFz7tq2erE0RWHrppcNII41UiA8HmXvmmWfC8ssv36ZPrMcSAkJACAiBLAKjZDfotxAQAs2NQP/+/cNaa60VDj744DDGGGN4MN/PPvss7Lzzzs1dcdVOCAgBISAE6oaACFzdoNSFhEDfIXDeeeeFjTfeOAwcODBMNtlk4bLLLgtzzz1331VAdxICQkAICIGGItAvsdLQGujmQkAI1IwA6bSQwqkIASEgBIRAZyEgAtdZ7a2nFQJCQAgIASEgBNoAATkxtEEj6hGEgBAQAkJACAiBzkJABK6z2ltPKwSEgBAQAkJACLQBAiJwbdCIegQhIASEgBAQAkKgsxAQgeus9tbTCgEhIASEgBAQAm2AgAhcGzSiHkEICAEhIASEgBDoLARE4DqrvfW0QkAICAEhIASEQBsgIALXBo2oRxACQkAICAEhIAQ6CwERuM5qbz2tEBACQkAICAEh0AYIiMC1QSPqEYSAEBACQkAICIHOQkAErrPaW08rBISAEBACQkAItAECInBt0Ih6BCEgBISAEBACQqCzEBCB66z21tMKASEgBISAEBACbYCACFwbNKIeQQgIASEgBISAEOgsBETgOqu99bRCQAgIASEgBIRAGyAwShs8gx5BCLQlAr///nv44osvwjvvvBO+++67omccbbTRwjTTTBMmm2yyMOqooxbt0w8hIASEgBBofwRE4Nq/jfWELYhAkiTh008/DTfddFM4+eSTw0svvRR+/fXXwpNMMcUUYeuttw6bbLJJmHHGGcMoo+hVLoCjL0JACAiBDkCgn00USQc8px5RCLQUAp999lk48MADwxVXXBG+/vrrknVffvnlw5AhQ8I888wTRhpJFhElgdIOISAEhECbIaBle5s1qB6nPRC4+eabwx133OHkbYwxxgh/+MMfQr9+/QoP99NPP4UffvjBjxkwYECYdtppw0QTTVTYry9CQAgIASHQ3giIwLV3++rpWhABhOLPPvts+Oqrr7z2q666alhyySXD2GOPXXiaJ598Mlx11VWuZo3HisAV4NEXISAEhEDbIyAC1/ZNrAdsFgQgZr/99lv45Zdfwsgjj+zOB2mpWqwn0rUPPvgg/Pjjj27btuaaa4Z11103IImLZfbZZw/333+/E7h33303/Oc//4m7unxiO/fzzz/79tFHH93v3eUgbRACQkAICIGWQkAErqWaS5VtVQQgbl9++WV4/vnnA4Rr4oknDnPMMUfAGQGP0nT597//HT766CMnXUjVJp100i6epnifjjvuuK5Whex98803Aa/VrB0c3qtI6J5++mknbosuumiYaaaZwlhjjZW+pb4LASEgBIRAiyEgAtdiDabqth4CkLc333wzDB06NFx44YUBBwUI1HLLLRd22mmnsNhii/lvpGRI0kaMGOGSNQgZRA0Sl/UynWSSSQJ/hBCB8KFSnXrqqZ0YIqlD2sf2Sy+9NBx77LFOCNk2wwwzhCOPPDKgloUAqggBISAEhEBrIjDyYCutWXXVWgg0PwKQtzfeeMNDgZx77rlu1wYZQ4368ssvhyeeeMLJ2ZRTTukk7Iwzzginn356eOutt1yiNv/884c11ljDiVn6abnGQw89FF544QWPEffwww/7tSCA448/vkv5jjrqqMD1Pv/884DqFAJHXLkHHnggTDfddP7HdhUhIASEgBBoPQQURqT12kw1bhEE0uQNyRvqTFSn008/ffj222+d2CF1Q8I2yyyzOKFDzUpBsobnKRK63XbbLUw++eRdnvqss84KJ510kkv3sJuDoFGwj/v+++/Dhx9+6LHjxhlnnLD00kuH999/P7z44ouBY5Hs/f3vfw+rr766JHFdkNUGISAEhEDzIyAJXPO3kWrYgghAplCbQrDS5G3LLbcMRx99dFhooYVclfrxxx97qJD33nvPiRXqT0jefPPNFzbeeOOw+eabu2o0a9sGJNiyYR835phj+rkQRv4++eQTvybnQNT22muvgDSOcCPPPfdciPd88MEHXaVKIGBlc2jBTqYqCwEh0NEISALX0c2vh+8tBAi+C3njDwcDpGxbbLFF2HnnnT0FFvd97bXXwtlnn+3BerGLg7hhF7f++uuHhRdeOIw33ngVVQ9bOcgiseOuvvpql7Ih2Zt11lnDHnvsEdZZZ52CowTODLvvvnt47LHH3Mu1f//+4ZprrvFAwHjGqggBISAEhEBrICAC1xrtpFq2GALYtmFeSiosiNm2224bdtllly6qUNSq2LLhKbrUUkuFueee21WntT4u0re7777bJXDLLrusS9iyxIy0XKThgsQRYoRUXVtttZU8U2sFXecJASEgBBqAgLxQGwC6btlZCKDiJFwItmjZQnDeFVZYwf+y+2r5jcoU1Wu5gpMD90XFGv/KHa99QkAICAEh0HwIKHli87WJatQGCEwzzTTuTABRwnng2muvDY8++mhRQvpGPCYSv3/84x8Br1XUrKTgWmCBBdxLtRH10T2FgBAQAkKgNgRE4GrDTWcJgbIIIAlbbbXVPMYbUi4IEzHZCA9STcG+jdAg2MjhVQoZJMgvNnYQsGoKDg533nlnOO+889wjljhw22+/fZhzzjm7xJmr5ro6VggIASEgBPoeAalQ+x5z3bFDEMDTFHUmHqbEfLvlllvCzDPP7PZwqDHLFcjZp59+6oQNZweIH4F5iR+HKpa4cVyLkCRTTTVVQSVa7pqvv/56OP74490LFa/TlVZaKWy22WZFOVbLna99QkAICAEh0DwIiMA1T1uoJm2GAHHcIEkxfRYpr5CALbHEEi6ZK/W4SNueeeYZd4DgeAIBkxc1XcihinPEggsu6JI+4rwRDqRcYN4bbrjB64LjAlkb9txzTw9Dkr6uvgsBISAEhEBrICAC1xrtpFq2KAJI2nBgIL7bDz/84LlLY8Dd7COhLkVFevHFF4fzzz/fiRtkC1I24YQTuncq6likcATqRSJH6JB//etfThTxdCXXKXZ3eQXJXcy7yjGEKeF6KkJACAgBIdB6CIjAtV6bqcYthABqS5wXIFuQOfKeksQ+r7z99tueHeGyyy7zlFd4r5Lyinhus802m6tKIWBka+C6ZFWIqtWrrrrK02ftt99+YcUVV8wNCUJKLuzwCOCLWva2227zmHRKbJ/XGtomBISAEGhuBETgmrt9VLsWRgBJGUnmUYdSiPFGrLc8+zfitxGPLZI3EtVz7EYbbeSfqEvT0jLSYWFXRxDe6667Lrzyyise123IkCEugUOlmlWnIgkkSDAx5yCBeMbiaIEnqjIxtHBHU9WFgBDoSASUSqsjm10P3RsIYKeG/Rqpqt59911PWwW5euSRR9zLEwnYuuuu20U6hpqUdFvkNoXIQd5IoXXggQcWVKLYvKULyezJj7r44ou7QwOSOJweuC/1mHfeeT37Q/Y8Um9hC4dEkMT2kElCi5CbFc9W6oLkL00W0/fVdyEgBIRAbyPwzjvvhBdeeMEdwBibJphggpAOSM44y0KUcE31LgRhp+TF7az3vXp6PUngeoqgzhcChgBkiFAh9957b3j11Vd94CHcB9vxKMVbdK655nLHgyxgHIfkjU/I01/+8hdPYI+jQXcFlSqpsrCrO/zww33Qwy6OLAx4pxIqJF2QtuH4QOotUnwdcsghfgzEjsEQKSHXIzYc184SwPS19F0ICAEh0BsIEKvy2GOPLVwaDQFpCPGiZ3F5++23B/JKo4mod1lrrbXcwQsnLyIHMA4ynjZjEYFrxlZRnVoKAeK0IWljcCFNVSwMNLz8rB4hRNi+ZQkRxOupp55ymzbitHEc0jfIVzUF6d6IESOcBLI6HT58eECNmiVw1GmVVVYJ5ETFYQLHCla4qFRRw+L1io0cAyj1zaphq6mTjhUCQkAI1IoA4ya2vmgW/vnPf4YTTzzRNQ5oMfqqYFuMjbAIXF8hrvsIgT5GAFH/lVde6eSN0CGoQBl8UE+i5iRh/PLLL+8SuGzVIH/PPfecS+oge9iozTDDDF2IXva87G9WqOutt56vTFGPPv744+7gMNNMM3Wxb4PAQRxRQaByjZJCVL/84XRxwQUXhD322MPrnr2XfgsBISAEehsBFrt43/NHXmlshBkr+5LAnX322b39mD26viRwPYJPJwuB4ISHhPQUgusOHDgwLLnkkm6bxuCDWjQreYu4ff75567ORM0K2Zpvvvlqtr3AWxWpGYQS1S2epni9Zp0mWFFusMEG/sd9sX2DuF1yySXh9NNPd4ncueeeGzbccEMRuNhQ+hQCQqAhCLDIRQLHGIqHfV5Bi3DQQQd5SCXMQfjOGMzYhknIDjvsEC6//PKAbR3By3fZZZfCZTAn4Xg0GJiYoI2IBRMTQkDtv//+rtUYNmyYq25POOEE347ZysILL+yHE0Xg+uuvD3fddZeHeWIj+7FT7q0iAtdbyOq6HYMATgDYk1GwW2PgiC91dyBAtPijYCcH4avVgQAJHgQSgkadCBycDQCcrQ/nIDHkD/s3flN4HlS6KkJACAiBRiDw1VdfhVlmmcVJFwtNFpfEucwrmJCwGMZODs0CGg9MSTjv7rvvdpvkww47zLUku+66q18XMoh6low5aEogeYx5999/f+EWkLpohoLJyY033uhj5G677eakcuWVV/ZrcMKAAQPCn/70Jx/Hhw4dGo466igPrl64WC98EYHrBVB1yc5FgGC8/FVaIFiRZCEpiwSq0vOzxxGcN4YEgcThVVppYfCKQYZLSQwrvZaOEwJCQAj0BAHMUbbbbjsfHzH1OOCAAzwkE9K4dMFr9F5zHuNv/vnnD5tssolL4m666aaC7Rr2c6uuuqqfhnaBcyBwZ555ZoAoYrvMIpZy3HHH+WfePxbHV1xxhS+ySWeIhgMJHtoO7ImxmWMBDtljO8SwN4sIXG+iq2t3JALVkB9c46N7PHHjqiF/eeCy4owkjFAj1dSFY6s5Pu/+2iYEhIAQqAcCELi99tqrcCnMPpZbbrmw0047FbbxBaJEOfTQQwvjF3EzyUEdC2NhLKQc/PDDD/0nThKLLLJIgbzFY0p9Qs6ihgSTFwo2xLPPPruHZoIcooXBEQy1a2+X/z1Vb99J1xcCbYoApIeXGvKFFIuVGO7tSMLiy17q0Yk1FOMNMaiQIqvWwv25BvenMIh150XKORBHpIDYkUQCSd1F5mptCZ0nBIRAvRFgPKOgIk2Pq2SpoRx88MEFiZtvsH+oUbMlPa5Btgi7xLgdF9LZ40v9jtdhwYz5C6GgIJmTTTaZB2DPEs1S1+nJdhG4nqCnc4WAIYDRLC8tKzFE/bfeeqsbwrINlSZ5R7Ftg6jFlz4Cx6CEKJ4BiaT1GNkyIHVHvOL56U+8T8nOgOqU1SuDCg4U2YJalbqiOsDWjeDBED9CiETyN8888xSIZfZ8/RYCQkAI9DYCLCwJOo5WAXteQhvh3b/CCiv4WBXvj+MWMTZxTNhnn338OzE5Gcu22GKLeFju56BBgzz8E04MRADA3q0as5P0RY844giXDp5yyildxvn0cfX8LgJXTzR1rY5EADsIBhW8nAjiSzw4bCUgZ8RzIywI9hb8TTTRREUYMSD98Y9/dJIHAYNE4YmKN1Q1BckZSe2x5WDAQ6SPQwNELltQG2AQHGPBYfyLp1dUvVJHQohUW4fsffRbCAgBIVArAmgEcE6gEGScGJnEt2RhnC4sfgm4C1lD6gVxY/w76aST0oflfmeMJCf0vvvu68dHe+Tcg8tsZPzF/g6SiX3dn//8Z/dWRV3bm6WfDdpJb95A1xYC7Y4AKzakbqeddpqL95GAoUJlBRlfL7xSsdHAaylbHnvsMQ89AvlDpE8E8tVXX71Lyq3seenfpNJi9Yk6AALHQMaglBcQmJhKuMEzQCIRRF2KxA93eRwp1lxzTT+3Jx6x6brpuxAQAkKgLxBg7EMdyljWl4WQSwRC33HHHV0DQlxQHBtwZshbRNerbpLA1QtJXadjEcBAFgkcyeLvu+8+l8Ih8sf+ApUqKkqkXsRnW2aZZbq80KwWkeKhxsTwllUc5In4QUjyyhVWfqhDOYd7s/rE9Z4k9dGrKn0+KlMkdaw0IW6oHyCN8Q/VKRkcol1e+lx9FwJCQAg0MwI99eKv9dmI/cbCeNttt/VLIIFDk/LMM8+4k0St1+3uPBG47hDSfiFQAQIMHIjQ+UMih3QLYkVGA6J5E+sNlSU2bqhM0wUbOcT/kDyyIBCH6Mgjj/TVHCQOIpZdxbHKRNKH3Rz3uPjiiz0OEsSPazF45A1mZGhgpYh0EC8q7DWwH8FWL2ufl66jvgsBISAEhEA+Amg7cKK45pprfNGNVoVYdMSF680iAteb6OraHYkAEjls2/jD7R2X8gceeMBznkKgsIlD+pUuELXtt9/eJWh4WUHiIHtI0hgISMcF0cPeAzUBhPDJJ58MV199tdtcYMMGedtoo438L3pspe+BdA47PZwXIGsxvVc2U0P6HH0XAkJACAiB8ghA4AgiTEYebIqJL4dzWm8X2cD1NsK6fkcjAFnCNo5AkqgvsW1jpYaELFuQquE9dd5557lNHRI8VKSoM/F0hRBCDiFrqGVxekDah3Ru3nnndTs68gQiscuTpkH2iCCOehcHBew0kBhW6z6frbd+CwEhIASEQN8jIAlc32OuO3YQAtiwQaj4RGqGswF2cXkEDiK1xBJLuKcqtmmkgEEaR55SVKXpAkHD5R1pHvZzOEeQwiumfUkfG7+jOkW1SyG8CHXKI3rxeH0KASEgBIRA8yIgAte8baOatQECxGVDhQp5wyYNt/LuwnMQBw63eYgZBI7QIEjjYn5SJG6oS7nWnHPO6XYWOFCko43nQcf1uC51wjvq0ksv9UjnSPZUhIAQEAJCoLUQEIFrrfZSbVsIASRnqC1vv/12z7CAmnOdddYJs846a7dPgb0boUdQcX755ZeuLkV6hpoVAgfpQrKHu3ylUjQSQRO4csiQIW6ncf7557sDA2FDsk4S3VZQBwgBISAEhEBDERCBayj8unm7IoCjAcmV8UrCXg2DVsgbTg3VxChCqgZR46+nhVhvm266aXjuuefc/g0pHAF95557bveMTaen6em9dL4QEAJCQAj0LgIj9e7ldXUh0JkI4GCA6hS7M+zNiBOHgwHOCI0sk08+uaecIfYcNnePPPKIe8dCOFWEgBAQAkKgdRCQBK512ko1bSEE8BTFAxViBIHD7g07tWwhUwOqVqJ2EyoESV02xEj2nHK/v/766/D88897oF7iuxFOJCtZg8QR943txINDRYtqVkUICAEhIARaBwERuNZpK9W0hRCAIJHGCk9PyBwepdi+rbLKKk6eeBSyIZCdATUr4UMIrLvUUkuFhRZayMlcpQntIYFI/F588UWPH0emBWK+cS2kfqhIY6gQ1KaENcExghAkEDzyAeYF/W0huFVVISAEhEDHISAC13FNrgfuCwRwMkBtOmLECLeFI8NCzIsKiUMyR+orAutiK0fwxxjwl+jd/OGNihNDnhSNZ+AaSNsgY6Rseeqpp/x+SNSIHwehQ4W7zTbbOJnjHti8nXvuuZ7mi/hyAwcO9Bhy3Xmw9gVmuocQEAJCQAhUjoAC+VaOlY4UAlUh8P333wdy5J166qmeiQE1JdK1tdde271Shw8f7gQMSdhEE01UULmi2uT39NNP72mxNthgAw8bkr059mvkQOUTyRpSNwoqUqR7eK1yrUUWWSSst956ntnhoosuKpC3zTbbzAP7Eo4kq2bN3qvW3xDJ3rp2rXXSeUJACAiBdkBAErh2aEU9Q1MigPqUpMaxkE6LVFqkyML2DO9UjllwwQWd1BHgl/RYSM2IG/fZZ585cVtmmWVyCRyq0jvuuMNt6PBsJYk96lKCAWMLd9NNN3kcOdJyEUCYbeRPjZK3XXfd1WPJ1ZNgQSKRKN58883hlltu8Tpkc79GPPQpBISAEBACtSMgAlc7djqzDREYPHiwqy8hXsRi62nhGssuu6xfBjs0yBTSMmK34dSAmhUJGyQNgoVNHEQPNSdED+JFAF/Ur+l4b0jz3nzzTZfkcd1VV13V/yCDSNR++umnMM8883iwXogU96QQjmTDDTd0yRtZHOpB3nDA4B78Qd6Q/lFw3sC+TkUICAEhIATqj4AIXP0x1RVbGIHHHnssHHbYYW7UjyRrpZVW8j88OmstSNkgcUjJcGQgLRaBc1GnYg+H5AzPU/aj/uSYSPSQwsWcp2nvVCRpH374oRM1zkNFmg7Iy/VXW201d4bAOeLVV191soZt3frrr+/bayVvSNnuueeeAmnLpvmKOCENrPUe8Rr6FAJCQAgIgXwERODycdHWDkWAHKVIknAQwHOUv3333dc9SiOZIxjv+OOPXxVCkLgBAwYE4q+RFgvSNfXUUxc8UtMXQ0oGkcMLFekbx1OfNIFDxYpTAqpYQo9wrawnKZI5sj8QwgSyh6MCqbSoS7UFAhilbDhfRClbuetwbxUhIASEgBDoHQRE4HoHV121RRHISzLPo3zwwQfh7LPP9j+IEI4BEDqSyHNOWr1Z6tEhYIQW4a9cgYhxDFI0QpDgXZole6haP/30U1etooqNcd3yrkveVP5qKRdeeGFArYy6ttoiAlctYjpeCAgBIVA5AiJwlWOlIzsAgUpIB16j2Knxd9BBB3l2hRVXXNEJHerJtKSsVshQo+JsgKQNZwRCgqTjwkHeCABMwUmAY3ujbLTRRuHpp58OJ554YtWXzwtcXPVFdIIQEAJCQAjkIqAwIrmwaGOnIoCzAOpRVJfVFFSiZ5xxRth8882rOa3ksXiiHnDAAeHWW291KVypAwk1csQRR7j9Wz2cLkrd57rrrguDBg1yR4tSx+RtJx4epDj9R3y7epDcvPtpmxAQAkKgUxAQgeuUltZzlkQAb0+kTATD5fO2225z54CSJ2R24IRw1VVXhZ44OmQu6amt8OjkuoTkQIWbTneFHdvCCy/s4UdwYJhssskqUuNm71PNb9SoSBhR6fakoCKeY445ikgdHrOogVWEgBAQAkKgMgRE4CrDSUe1AQIQoJdffrmIrEHYyFxQayEEyLBhw3pFhUkQXGLDPfTQQ+7IwO9YkPiRpQHSCJmrxAYvntuTT8KT7L777mHo0KFlLwOhxPGiGmyRJqYldXzHAUNFCAgBISAEuiIgAtcVE21pAwQIdfHss886WYvSteeee66QraCnj4gU6fjjjw+77LJLTy/VkudfeumlYdttt/XAwHkPcMwxx4T99tvPCSiSzfQfse0qLThfIJ3DUSSSO1SwSv1VKYI6TggIgXZFQASuXVu2g56LOGlR/QlZ44+wF2mVYzk4sNOKBIFPAtCSBL5UQSp05ZVXehy3Usd0wnakmahUyceaLdju4diRVwhYnCZ0fMdJA4ldJQVnjjwV7LjjjlvJ6TpGCAgBIdAWCIjAtUUzds5DYK+WJWvvvfdexQAQBw2SliZs/fv3Lzqf2Go4BOQRCgLvElqj1rAcRTdqgx+EOdlxxx3D+eefX/Q0eMiiRq20gDkkLk3sRowYUbEKFhVyngqW9lYRAkJACLQjAiJw7diqbfBMPbVXI4gtzgWRrEXCRpL4SgoZC5DkxcL1Dj/88LD//vv3mb1ZvHcrfJ5zzjlh5513dhU1QYgJPlyPEgl7mti9/fbbFV8aoh1Vr/FTKtiK4dOBQkAINDECInBN3DidUjWkONinRfUnk3U19moEvMWYP03W+I0qtNay5ZZbej5SzkeShM0X+UpVSiOAzSEqVXKskrWht8pXX30VkM6lSZ1UsL2Ftq4rBIRAsyIgAtesLdOm9aq3vRqkjaC3SMjqWU499dSw6667hqWXXtrJW7sGpcXZA5s0pGb1KN9++214+OGHwworrFCPy1V8jawKlsUAJA+yV0lBBQvxjFK6+EkGDBUhIASEQDMiIALXjK3SJnXKxldjUq23vVpvQfXggw+GG264Ifztb38rSw4vvvji8Prrr3sMM+KYlfpLZ1God52RYJKxgZAdpf4gMnn7CAtCSrDelJjV+3mruR7q1rSkju/0y0oLKvdI5uInCwZ5wVaKoI4TAkKgtxAQgestZDvouo22V2sk1Ndee21YZ511uq0CBC6P3C2xxBJhjz326Pb8cgcMHz7cA/qm48SVOz6778YbbwyrrrpqdnPb/obMZkkdKlikeJUU2nLOOecsInaEOumtdGaV1EnHCAEh0HkIiMB1Xpv36Il7aq9GAFrs05BmRJu1ntqr9eiBengy5HXGGWesSqoTbznppJOGRx99NEw33XRxU82fJ598sgfYrfYCeG4iQRxppJGqPbWtjsfjOM8LticqWPr3VFNN1VY46WGEgBBoHgRE4JqnLZquJq1ir9Zo4I499lgPWltNPXC8uPvuu8Oiiy5azWllj8UL9PTTTy97THbncccdF/bee+/sZv3+fwR6qoKdeOKJiyR1LFz++Mc/SgWrHiYEhECPERCB6zGE7XGBGK4heoK2kr1ao1sA+zOM3XEIqLRccsklYeONN6708IqOQxq4xhpreO7USk5AGvr+++8rpl0lYKWOwZYwq4J96aWXKlbBQt6zKti5555bKtgUxvoqBIRA9wiIwHWPUVsd0cn2ar3ZkIMGDeoSzLbU/QYPHhwOPfTQUrt7tP27774LSy65pBOM7i6E+vTcc891T9vujtX+8giggn3hhReKiB1esHj4VlLwgkUVHx0l4qdUsJWgp2OEQGciIALXxu2eZ69GrK4ff/yxoqdGQoOkINqq8dnK9moVPXSVBzFpE3LkggsuqEgCh9QN6Vtvlg8++CAsvPDCgc9KCm280047hU033dQzUFRyjo6pDAHyvmalde+++25lJ9tReSrY3gibU3GFdKAQEAJNg4AIXNM0Rc8qInu1nuFXzdl4exJi5JRTTnE7tkrPxd7tnnvuCb0ZUiTWBRX4UkstVTLZfDwu/UkuUSSJpMbCTkuldxBA5Z4NRFyLCpYFVZTUoYIl/ZuKEBACnYOACFwLtnUrx1drQbgLVcb26eyzz3ZHAYzbswWVJBKXvIKnKR6neJ72VSE8yFprrRVQm2cLjhc8A5JD1K7pgjpvueWW89RYq622Wsd7qKax6a3vxOPLesEiufvmm28quiVtNtNMMxUIXSR2U045ZUXn6yAhIARaDwERuCZuM9mrNUfjRDUpSexRS6cLAV3XXnvtsMsuu4TFFlssN6QIkq2HHnoozDHHHOlT++R7zCiRvhmpwVDjjTbaaE4QIHF4r7788svpw/x7//79ww477BC22morV+d1OUAbeg2BJEl8QZBVwVYTDHuSSSbpQuqQrtY7c0mvgaALCwEhUBIBEbiS0PTtjjx7tUbnA+1bBJrrbt2pSbFN2mabbVzdmE63lA0pwkSJJIxsB40qu+22m6t74/0POuigcMQRR8Sfhc+77rornHbaaa4ezkrt8JzccMMN3VZuwQUXLJyjL32PACrYLKlDBfvrr79WVJlo2xqldHxKBVsRdDpICDQVAiJwDWgO2as1APQKb9mdmpTJjhypOCNAarIlG1IECRjx2RpZIKNrrrmmE0kkhqh506QzWzekc0OHDg3Dhg0Ln332WXZ3WGihhfyZNthggz6x5+tSAW3oggAq2Dwv2J6oYLGxa9ccwF0A1AYh0IIIiMD1cqPJXq2XAa7T5StVkxKio7uCZA7yA3GDwDVD+f777z28CAnbr7rqqoqqBCm44oorXCr32GOPdTkH9dzWW28dtt9++zDttNN22a8NjUUgq4LFsQXJHbH/Ki3YbKYldXyfZZZZpIKtFEAdJwR6EQERuDqBK3u1OgHZh5epVU3aXRVRfe+3336uimwmW6MPP/ww8LfAAgt09whd9j/xxBNO5C6//PIuYWh4xtVXX93Vqzg/qDQ3AlEDkFbDYv9YjQo2psOL5A4V7FhjjdXcD67aCYE2Q0AEroYGzbNXU3y1GoBs0Ck9VZNWUm2kV30RLqSSutTzmM8//9w9cc8444zc/K/EKCMMyeabbx5w3lBpDQTor88//3yRbR2hTr799tuKHoBcunlesFLBVgSfDhICNSEgAtcNbHG1mk4x9eqrr+aGZsi71AQTTOCBcFmpxoC4CsSZh1Tvb6unmrT3a9vcd0B6iXMGTg933nlnQF2XLsQkGzhwoEvlZp999vQufW8RBGjTN998s4jUSQXbIo2nanYEAiJwqWZWPtAUGG3ytbfUpG0CT10e45VXXgn/+Mc/PJVYXuqoZZZZxokcjhQ4Uai0NgJxUVurCnbMMcf0jC5R/conKlmpYFu7X6j2fY9ARxK4tL1aNOzlE9VaJQWbHwx5o0QtRkSfaKKJKjldx/QBAn2hJu2Dx2ipWxAQ+KKLLvKYcqjjsgXP1+222y5su+22fRrQOFsP/a4/AqTnQ8Idx1PIHWYl1ahgZ5555i4OE5NPPnn9K6srCoE2QaDtCdwPP/zgA0lUgTKwKL5am/TenMeQmjQHlAZsuu+++1y9et1113UxjieA8Prrr+9SOdKLqbQnAqhg33jjjS4q2Epz9IIKQafTkjq+s3jG5k5FCHQ6Am1F4KJoP5I1PmWv1v5dXGrS5m1jJuszzzwznHXWWeHjjz/uUtE//elPTuSIq0eAWZX2RwBHmLT6le94wWaDR5dCYpxxxglcg4WAihDoZARalsApvlond9v/PrvUpK3TB3755RePP4fTA2nFsmXCCScMW265pXuwklNWpbMQQAWb5wWbzdMLKqQCy0v71lmI6WmFQAhNT+DS9mqs1KJ0TfZqndt9pSZt7bbnPYbIXXLJJQETh3RBNbbyyit7EOQVV1wxkKRdpTMRQAX7+uuvd5HWEUz7sssu60xQ9NRCIIVAUxG4vPhqsldLtVYHf5WatP0an0XYOeecE4gph61UtkDydtppp+xm/e5wBJDmjjrqqB2Ogh5fCDSRBI78i6T5qdQOQvHVOqv7Pv74456DM/vUGDWXy02aPV6/mw8ByPmtt97q3qu33HKLx5Qjzywpn+TZ3XztpRoJASHQHAg0jQQOcfn4448f8pIvTzPNNIWQHUzYhO3o379/cyCoWvQZAgsvvHAgJyexxNZee+2wyy67eH7PPquAbtTrCCCJI6YcpO7EE0/s9fvpBkJACAiBVkWgaQgcABLw85NPPimQNcVXa9Vu1Tv1vvbaawOSOFI1EVNMRQgIASEgBIRApyLQVASOVbfi+3RqV9RzCwEhIASEgBAQApUi0FQErtJK6zghIASEgBAQAkJACHQyAgpn3cmtr2cXAkJACAgBISAEWhIBEbiWbDZVWggIASEgBISAEOhkBETgOrn19exCQAgIASEgBIRASyIgAteSzaZKCwEhIASEgBAQAp2MQJ8SuIcffji89957XfAmBhzpde65557w888/d9mf3vDKK68Ewkm8+OKL6c36LgSEgBAQAkJACAiBjkGgTwncn//8Z0+dk0b3119/Deutt57HfmP/AgssEErlOd10003DrLPOGtZZZ50wxxxzKM1OGsgm/V4P0s6jvfbaa54Ht0kfU9WqEYFvv/023Hbbbb54I5VeuUKMyBtuuCH861//CqRTUulMBFjwP/roo2H48OHhww8/LAtCNf2r7IW0s24IIIS57777/I/5oTfe5bfeeis8++yzdatz017IXoY+K5YeJxk8eHDR/SzaejL66KMnlkInef7555PJJ588GTRoUNEx8ccss8ySWBLj5Kmnnko22WSTxEBNbrzxxrhbn02IQF6b2wubGAn39qMN55prruSLL74oW/vNNtssMdJe9hjtbC0ELChzMsUUUxT6wfTTT+9jQN5TXHrppclYY41VOJY+89FHH+Udqm1tjICR/GTZZZct9IPRRhstOfPMM3OfuJr+lXsBbewVBLbYYotC+zH+TzjhhMlWW22VfP3113W73x577JFY5h6/3jvvvJOcdNJJNV375ptvTu68886azu2Lk/pUApfHYklkvfrqq4eVVlrJpWrbbrttMJIWvvrqqy6HG5Bhww03dGndqaeeGuzlDQ888ECX47ShuREgSflNN90UyHtppD189tlnYc8992zuSqt2dUdg4MCBYeKJJw7PPfdcYCXO+2yDe5f7GLkPW265ZVh66aUDq3f6zuuvvx7222+/LsdqQ3sjcMwxx7i09qqrrnJzHOaDnXfeObz66qtdHrzS/tXlRG3odQRmnHHGYITN23DIkCHh7LPPDoceemiv3HfEiBEBvlCJpI9kAulCP7v++uvTm5rqe0MJHOJtXjxbTRdAQTX6448/+sRe2Pj/X8iJGsu4447rOTH5VGktBKoh7a31ZKptpQhAyl566aVgK+8w55xzhkUWWSRsv/32niqNJPbp8vLLL4cffvjBc9+aFD6sssoqYfnllw8MzCqdhcCDDz4YFl100bDuuut6Or2jjjoqYIZjmpgiIKrpX0Un6kefIEDGJeZuUiJuvfXWYcEFF3S1eG/cHAERC75RRx217OW/++67LmZZEMtTTjml7HmN3NlQAhcHagbwWCBwlLgvbs9+3n777QGbmZVXXjm7S7+bGIFqSXsTP4qq1gMExhtvvGAq0cB7zARMMXW7f3788cf+Gf8xJrCPwRT7WBye7r333mAqkniIPjsEgammmsrbP9q+MSmPPPLIIdtnqulfHQJd0z4mbYkmJgpyWKAxLpxwwgmeH/2aa67xurNgW2uttcLcc8/tEvnPP/+88Ew//fRTOOSQQ9yGnsUg2p1Y7rrrLtfwxd/ff/992Hffff3Y+eefPxx99NGBeQkt4AUXXBBMRV/gFVyT/bFw7t/+9jevBxLeWDf2I1HkXKR266+/flhooYVc8hfPfeyxx1yDuMQSSwQzJfPj475aPxtK4FCZUFhdx4L0jWJ2cXFTl09EoajcVl111TDvvPN22a8NzYtAJOa1kPbmfSrVrFoEmHQPO+ywYDYmoX///gHJ2g477OCXmWGGGYoux2R81lln+cA42WSTuQmF2cqG4447rug4/Wh/BP7617/6Q9JHmCCnnHJKXwBk+0w1/av9UWu+J2QewGmRRdi0007rphS77babV5RoFKjKITxmHxcgYGbHFhZbbDHnBbvvvrtr7pDcxfKXv/zFyRJmFqjUxx577LjLyT3XjGW11VYLaIEgehx/xx13hCeffDKMOeaYft6KK67oEn6OhzTGiBfffPONjz1mtx/QBpoNrjtg7rPPPn5pImjcfffd4YADDnBHSwjhrrvu6k5aHLDNNtu4dhHixwL03Xff9fN68m+Unpzc03N5+Siw71iwh6EgWi1VGPhh39jEqbQWArWS9tZ6StW2EgT22msvJ26slmeaaaZw/PHH+4DNoJ0uqDYYcNm+xhprBFSqTzzxhNvCbbzxxulD9b3NEZhtttncG33YsGGhX79+TuCwUVpqqaW6PHml/avLidrQ6wggoJlnnnmcuCGMQXKFPWwsqFexhaeYM0CAuLN4++c//+kS1/nmmy8svvjiAYkYocmQhA0dOjRst912fo45OoaHHnrIv6f/sQ3ydPXVVzvJSu/Dux1iiXQur2Crx/4PPvggTDDBBH4I6lUIJWQyjlsQPIRLlHPPPdfHKkghWoRPP/3UpXvY+tejNJTAjTHGGM7An3766cKzwIQnnXRSb9zCxtQXQg7AzrF5QJyu0loI1EraW+spVdtKEcA+hb8rrrjCV7SoJ7IFKR0DL/ZPrMIpK6ywQjj88MODCFwWrfb/PfPMMwcmU9TpSN4wo5l99tlzH7yS/pV7ojb2KgKTTDJJgOiUKkivYoGov/nmm4GFHCrUWJDevfDCC06K2MbirrvC4m+UUUZxO9rujs3uh5tQr0je2G/REZzAQRiXW245P4Xrx4KzRlT3Q+YImcYiBI1CPcauhqpQeUjYK6TsyCOPdAYNw0YEiqQGvTbiSGKAUWC/xIJDB06cF1bl/PFdpTUQqIW0t8aTqZa1IvDGG28E1BCoNHi/s4UVL4UVeCxM3HF73KbPzkEAMxqcXlBbHXvssWUfvLv+VfZk7WwIAswT6QLpwfEhxoxEJcofavSorYPMdVcwvcDm9pFHHulyKNfPeqGmD7KQRy5N++233wqbLVSNSwjTY1Nhp32BfMbCIgP7XSSOqHzzJITx2Eo/G07gNtpoIzcS/Pvf/x4QeSOKPPDAA73+6JhRq9x///1O5mCvqE6R2O24446Fv7QEr9IH13GNQ6Aa0t64WurOfYEAdi54FbJYu/LKKwueYphVHHzwwT6gYtSM+oFB75xzznFjZYyNK1lx98Uz6B59iwDG4kjd6C8WA869mGMN6BcYkcdSqn/F/fpsDQTWXnttDy1m8V/ddArytssuu7idHKpYFn84BhBWjMgWeWHIeFJs3rC5RUiEGp7+gbMEBRUuZB9yh4YPaV26sGBAmkYdsHXDqQobN+6PY0K5gqoXbsM1N9hgAxdQ1UXw1BfB5iq5h7HixAbxLoeag0OXbdrQOgjYxNsleDO1N8+eZPzxx0/McDQxMp7YqsYfyjp1Yp5lib0cRQ+pQL5FcLTNDyNkiXmCJW+//XbRMxHgm75jNiO+/dZbb03MTs4DgNI/CORtg3TROfrRGQiYOi0x1VRy3XXXdXlgU7EnpuYqbC/VvwoH6EufI0AgX1ODl7wv7/15553XZb/ZuXnQd5Nq+RxhC7hCMG+CeltMwGSiiSbyMcLIVCGQ70UXXZSYRq9wPbOzT2zRmJjEzf9Muuf7TDiU2MIgMRVoYuZZiZlsJNzDtAKFcy0TTNE4ZCrdxGzwfD9jFfdlrIplwIABPr/BbQhATd1JXMB5Jj2Oh9X82Y8zyzFH7RMCvYUAomj+omNDvA+eyEhcVDoDAfoAXoPZktcPWFkTfqS7mE7Za+l3eyFQqs+gWkUVlu5PpY5tL0Q652kI+cEYQDv3pBD9AvqD92m6IC3j+uUKUmCOSdu7lTs+7it1z7i/2k8RuGoR0/FCQAgIASEgBISAEGgwAj2jsA2uvG4vBISAEBACQkAICIFOREAErhNbXc8sBISAEBACQkAItDQCInAt3XyqvBAQAkJACAgBIdCJCIjAdWKr65mFgBAQAkJACAiBlkZABK6lm0+VFwJCQAgIASEgBDoRARG4Tmx1PbMQEAJCQAgIASHQ0giIwLV086nyQkAICAEhIASEQCciIALXia2uZxYCQkAICAEhIARaGgERuJZuPlVeCAgBISAEhIAQ6EQEROA6sdX1zEJACAgBISAEhEBLIyAC19LNp8oLASEgBISAEBACnYiACFwntrqeWQgIASEgBISAEGhpBETgWrr5VHkhIASEgBAQAkKgExEQgevEVtczCwEhIASEgBAQAi2NgAhcSzefKi8EhIAQEAJCQAh0IgIicJ3Y6npmISAEhIAQEAJCoKUREIFr6eZT5YWAEBACQkAICIFOREAErhNbXc8sBISAEBACQkAItDQCInAt3XztV/lXXnml/R5KT1QTAr/99ltN5+mk9kfgP//5T7j//vvb/0H1hEKgDAJNR+AuvfTS8Oabb5apsna1KwLPPvtsWG211YIm7nZt4eqe64ADDgjvv/9+dSfp6I5A4OSTTw777bdfRzyrHlIIlEKgqQjcL7/8Eg488MBw7LHHlqqvtrcxAkzYr7/+erjiiiva+Cn1aJUg8O677wYm6eOPP76Sw3VMByHw5Zdf+hzx8MMPh/vuu6+DnlyPKgSKEWgqAnfmmWeGt956K5x//vnh448/Lq6pfrU1Ag888EC46aab/BmPOeaYtn5WPVz3CAwePDj89NNPYdiwYeHf//539yfoiI5BYMiQIeGrr77y59VY0THNrgfNQaBpCNz3338fjjzySK/ijz/+GE488cSc6mpTuyKw//77Fx4NVerNN99c+K0vnYXASy+9FC644AJ/aMaFU045pbMA0NOWROCjjz4q6g+33npreOaZZ0oerx1CoJ0RaBoCd9JJJ4VPPvmkgPXQoUPD119/XfitL+2LAJI3JHDpcvTRR6d/6nsHIXDQQQcV2UGedtppASKnIgQOP/zw8MMPPxQBISlcERz60UEINAWB++KLL8Jxxx1XBPs333wTTj/99KJt+tF+CCRJ4naP2SeD0D344IPZzfrd5gg8/vjj4Zprril6SsYHzCtUOhsB7GNRqWfLVVdd5baz2e36LQTaHYGmIHCsoPKkbRgxo05VaV8E8DoeMWJE7gNKCpcLS1tvTKvS0w96wgknhJ9//jm9Sd87DIFDDjkk/Prrr12eGq/1rACgy0HaIATaEIF+JgFJGvlcH3zwQZh55pm7iMVjnVCf7LTTTvGnPtsIAbyOZ5tttvDGG2/kPlW/fv2c3M0111y5+7WxvRC48847w/LLL1/yoZC+bLXVViX3a0f7IsAib7755gulpqvRRx/dHeCmmGKK9gVBTyYEMgg0XAKXZ9OQriNhBPJWXelj9L01EWBCLkXeeCIGa9m3tGbb1lJrwsiUK4QX+v3338sdon1tigCS2VLkjUfGY1mOb23a+Hqskgg0VAL32muvhdlnn71bgnbhhReGTTfdtORDaEfrIUAk9ZlmmingVVaujDzyyIF+Mv3005c7TPtaHIGrr746rLfeet0+xZVXXlnRcd1eSAe0DAJkXFhqqaW6re8444wTiB84/vjjd3usDhAC7YBAQyVwBx98cLfkDZCJ+1Nu9dUODdFpz0BoiO7IG5jIvqX9ewZtjOdpJUUS2UpQaq9jStlFZp/y22+/leNbFhT9bmsEGiaBe/rpp8P8889fMTEbPnx4WH311du6MTrl4YikPsMMMxSCcXb33H/4wx/C22+/HSabbLLuDtX+FkTgnHPOqcq27fbbby9rK9eCEKjKJRC48cYbqxr3J5lkkvDOO++EMcYYo8QVtVkItA8CDZPAYe9SjVRNK+/26XTYMsVI6pU8FZ7IxAlUaT8EsF0i60I1RWNBNWi17rHYO3ZnF5l9us8++yycffbZ2c36LQTaEoGGSODIXzdgwICqAeW8Smwhqr6wTugzBFCbYvuGDVw1ZbzxxnP7lnHHHbea03RskyOA4fmee+5ZdS0fffTRsNBCC1V9nk5oHQQuvvjimmyf+/fv73HhRhlllNZ5WNVUCNSAQEMI3KBBg8K9997rcZ2I7ZT+KyeVW3nllZViqYZGbqZT3n//fQ8Nkm5zvqMque6667yqSy+9dOAve8zAgQPD4osv3kyPo7r0AAHe9W233TZ8+OGHRW2NxDWmRxpppJHC5JNPXrSffrHmmmt2Cfjbg6ro1CZEAJs2/rLjwN577x3uuecerzESugUXXLDLMfQPhRRpwkZVleqKQEOWKOedd17uQ2DrhEqFEtNqpV9e4oaptDYCU089deAvW4gHmCZwhx12WPYQ/W4zBIjzd9ZZZ3V5KtRgk046qW+faKKJAn0jW3B8UGlvBPAq5S9bxh577MKmhRdeOKyxxhqF3/oiBDoJgYYQuFIAQ9ZiwRiVAV6lMxBIt/1oo43WGQ+tp8xFoJK+QHgZlc5EoJL+0ZnI6Kk7DYGGOTFkgSZYb1SfYrsg8pZFqL1/a1Bu7/at5unUF6pBq/OOVf/ovDbXE+cj0DQETi9lfgN1ytaoOud5JYHrlFbPf06NBfm4aOt/EVD/UE8QAv9FQAROPaEpENCg3BTN0BSVUF9oimZo2kqofzRt06hifYyACFwfA67b5SOgQTkfl07cqr7Qia1e+TOrf1SOlY5sbwRE4Nq7fVvm6TQot0xT9XpF1Rd6HeKWvoH6R0s3nypfRwRE4OoIpi5VOwLpEDGygasdx3Y4UxN0O7Ri7z2D+kfvYasrtxYCInCt1V5tW1s5MbRt01b9YJqgq4aso05Q/+io5tbDlkFABK4MONrVdwhoUO47rJv9TuoLzd5Cja1fun+MPvroja2M7i4EGoiACFwDwdet/4dAelCWCvV/uHTiN/WFTmz1yp9Z/aNyrHRkeyMgAtfe7dsyT6dBuWWaqtcrqr7Q6xC39A3UP1q6+VT5OiIgAldHMHWp2hHQoFw7du12pvpCu7VofZ9H/aO+eOpqrYuACFzrtl1b1VyDcls1Z48eRn2hR/C1/clpj/VRRx217Z9XDygESiHQNAROXoilmqgztmvS7ox2ruQp1RcqQakzj1Hf6Mx211PnI9A0BE4vZn4DdcpWtX+ntHT3z6m+0D1GnXqE+kantryeOw8BEbg8VLStzxHQwNznkDftDdUXmrZpGl4x9Y2GN4Eq0EQIiMA1UWN0clU0MHdy6xc/u/pCMR769T8E1Df+h4W+CQEROPWBpkAgPTArOGdTNEnDKpHuC4oJ2LBmaMobq280ZbOoUg1CQASuQcDrtsUIaGAuxqOTf6kvdHLrl3929Y3y+GhvZyEgAtdZ7d20T6uBuWmbps8rpr7Q55C3zA3VN1qmqVTRPkBABK4PQNYtukdAA3P3GHXKEeoLndLS1T+n+kb1mOmM9kVABK5927Zlnuy3334Lv//+u9d35JFHDiON1DTdsmUwbKeKKiZkO7VmfZ9FBK6+eOpqrY3AKM1SfYyVJ5xwwsALOuaYYzZLtVSPPkCgX79+4ZJLLvG2h8ypdDYCM800U1hiiSW8P0wxxRSdDYaevgiBueeeO7z00kveN+TsVASNfnQgAv0SKx343HpkISAEhIAQEAJCQAi0LALSVbVs06niQkAICAEhIASEQKciIALXqS2v5xYCQkAICAEhIARaFgERuJZtOlVcCAgBISAEhIAQ6FQEROA6teX13EJACAgBISAEhEDLIsBijE4AAEAASURBVCAC17JNp4oLASEgBISAEBACnYqACFyntryeWwgIASEgBISAEGhZBETgWrbpVHEhIASEgBAQAkKgUxFoeCBfoq7fc8894bzzzgv//ve/i9qBIJ5bbrmlB/UcZZSGV7WobvrRMwTIvPDll1+G1157LXz44YeFTAxcddRRRw3TTjttIKDr2GOPHQj0q9LeCPzwww/h8ssvD0cccUT46quvCg9L288222zh0EMPDcsuu6z6QgGZzvry66+/hldeeSXce++94aOPPioaL8Ydd9yw5JJLhnnnnTeMNdZYnQWMnrajEWhoIN9vvvkm7LXXXuGGG24In3/+echG4Ye0TTDBBGHQoEE+gOvlbJ+++v7774c77rgjXHDBBeHhhx8O6fRJDMhM1gMHDvSBeaKJJmqfB9eTdEHg66+/Dptvvnm4/fbbw48//hiyscVJrTb++OOH7bffPhx++OGBdGsqnYMA2XnoG1dffbV/fvzxx10I3KKLLho22GCDsOGGG4rEdU7X0JOSiaFR5eyzz06MoJEJouzfOOOMk9x0002JrcIaVVXdt44IfP/998mRRx6ZTDnllGXb/Y9//GNCH+F4lfZF4Nhjjy3bD+L4YAu45K677mpfIPRkuQg8+eSTycorr5z84Q9/KNtPGC8eeuih5Jdffsm9jjYKgXZDoKF6SRuMw3fffecseq211gpzzjlnUSLzRx55JNx9993h22+/Dffff79LZbT6bv1FxyeffBJuu+02V52iIiUHLrlwY0GdhnoVlcktt9wSFl544TDHHHPE3fpsMwQYB2LZYostwjzzzBN/BpuMA/tvvfXWgBqNceDPf/5zYb++tD8C9913XxgxYoRLZ2eeeeYwzTTTFI0XqFRfffVVHy8wx5lrrrnc9KL9kdETdjoCDSNwqMxef/11V5tCynbaaaewwAILFKlHmLwfeOABH7hffvnlLirWTm+8Vn1+yNl7773n1Z9//vnD+uuvHyabbLLC47zxxhvhyiuvDLby9uPeffddEbgCOu31BZXpm2++6Q/FOLD33nuH2WefvfCQkLappprKCRx2k0zUKp2DAGY1LORYxKNKZ6G/4oorFhG0xx57LPz9738P77zzTnjuuefcHIOFoYoQaHcE6k7g/vOf/4Tnn3/eJWuzzDJLMDVZkVQtAorNWzRen3jiicP0008fxhtvvCIjZROJ+0qLazLIsxovVZj0WaVhJ4c9BHZUKn2LAKSctsLwnHbAGSFbGJAxUofEYeP4pz/9KWy88cYuhYvH0pYvvPBCeOqpp9w2MuvcEo/jkz5hKlb/HGOMMcKYY46Z29/S5+h77yOAZB0JOgQNko5DUl759NNPfRxgH+NAmsizjUkbhxb6Cn3nrbfeKiz62J8uEDwWehB/bOaWWmopH1PSx+h7cyDAe0vbM14ggae98rQr9CPmCcYWU6GGhRZayJ3aRh999MKDsB0nOAgc/SNtT1s46P+/fPHFF3497oUkj3FKTlJZlPS7VRCoK4FDonbIIYeEBx980AfuySef3CVrm2yySRfDUo6N3maQPAzVsy8SK29eboycMXqHxKFeYVCPhYGAl/eEE07wyZ5rzD333OGUU04pWsnH4/VZfwSYOBmMn376aSdeDK60ASrxrAMCgysDMgM3gycTNqQrXRjMOY9JG6KH0TKGzGk1q9kyBJxgWHE/8cQT3pdYBEDeZ5xxxtzJIH0Pfe89BJ599ll3SnjmmWf8JpA3nA+23nrrLjeFqEPAKRA1nJbShfd56qmn9v6A6h3JLZ7Ls846a/owXzCaLV04+eSTvV8wRkAczYbSVWpFB+tHQxGAROGUgDqcOYD2XW655fzdzUrOePdpd8Z5xopJJ520y8KQ7ZEAMk8wFjH3pOcJyD/E/sYbb3SSzz7mEhwfpptuOo0XDe0RunnNCNTLqM9UG25oai9gwdDUVjnupLDNNtskJlVJbKJPbLJNDjjggMRcvhN7ifxYE4vnGqrbS5uYe3jCdWwgT4wQJDvvvHNiruTu0PD2228n2267bWIrd99vIPj1TPKT2OCdmCSwXo+n65RAwAbGhHbAKcEIt+NvxMvbytQavo92xAHFBtbEyH1i6nI/rn///u6kQL9IF5PaJMccc0xig3JipC3ZaKONErOFTGzyTsw+LjEyl5haNRk6dGiyyCKLJLQ3bY9DjIWdSWwVLoeXNKB9+N1IW2JEzNsjvo/xc7PNNvM+QHX+9a9/eVvZ5Fs49i9/+UtuTW0Bl8w333x+HOMAzi9bbbVVYrZRCf3PJO+JeS0XrhPvx7FG9Hx/7oW1sc8RMGl6sttuu/lYwbhOW/H+0md22WWXxAi9OyHYAj8566yzknXWWScxzYwfx1yQN6YzHgwaNCixBaHPFSZ5TQ4++ODEzG8SWzAmtmBMbAGRmB1tYovLQj8xiX2yzDLLJCYpluNDn/cE3bAeCOCy3+MSyRsvBC8kL5zF8EqYyPltEpZk8cUXTxjAeVHxKo3kjf2HHXZYwqSdV5jsmcQ5jnN4SW0V5i+22cz5b/bxZ+J1J3N8Z3AQictDtH7bIGWQpcGDBycmKXUSzWAcvcUgVEy0EG4mW45bYoklElNve3vRfhZCpkuFIHTnnntuYhI1P47rMfjuuOOOybXXXpvceeedye67714gjOynj8X+YTHDvF7UT6XvEDAJbBF5Mymqt1t8P/k0+7bEbJiK3v+4/6STTsqtrElrvR/F4+In7zj9ySR83k/YzjbGmkkmmcS3icTlQtqQjZC3U089NTHpmI8VLPZN+lYYLxhD8DbdddddnZAzV/Be04a0LYszFot5hQVkXAwwBtH+LPhZ4K+xxhp+zzjnWFxBnyf4DaEbMGCASFweqNrW9Aj0mMBB3hiQI3ljcj7jjDMSU235SguyxsuXntjjQMukDEGDBGSlMBE5U5ckZtjsZCy9euJ7XMExge+zzz6Jic8TixXkZCLeAxLHqk6lvgjkkTcG4BVWWCFZbbXVChMobWOeYwmDZhxAaTcGWFOpeT/Jqxmr5/XWW8/JerrdGfwhdnFg5570P4v/lJjq1O8BcRSJy0O197ZB3symqECkaBfzCExMze2SlTgJ816m/+gLvKO0F8fmFcYGJK5//etf/di4MExfh++0+/HHH+9SF/NaTSCQbI8kzlS7eZfXtj5AAPJmZi0uEaU9WIjvueeeHh4KwhYJN+MFAoDYxvQPiNzaa6/tCzfIfF558cUXvX+YWrRACGl75qM4fvCJNJ9+ec455/g4Ql3YDol79NFHJYnLA1fbmhaBHhE4s2VI1lxzTZ9M48ty+umnJ2xHtWH2S4kFavVVdyRbrLrM6zC54oornLhxDMeWKuwzOwknZ2a/4CvxODBzT8gB9+AYCpI882AsInGs0rmPSn0QYEL94IMPXHLKpMkgyIRtISAS8xhLUH8w2c4wwwwF6SltBck3exMfLFFpmPeYq0PzaoWqFHUrx7GCJsYTA3skgUhlIXKoXSAPFkogGTJkiN+T+lAvJLuoT0otDvLuq23VI4BqHIk7bcwffSEbr+2qq64qks6xsEM6YkGcq1J3oy6D3GOWEVVr9Amk72yPhQXGzTffXCBxHANRZGxS6VsEeJchTCzYmQcgb3vssUdiDic+XrPwZtFvDk0FcwiIHOM2kjUk7ggKzKGhZMUx0+BdJ2Yc0ntzjHJSyFhAn2Rx8be//S156aWXvL+hlmfeYDyKJI6FYDT1KXkj7RACTYRAjwjc8OHDXTTNC4Id2j/+8Q8fINMTJismXioGa14qSJgZpbotU7U4YM/AAIz9DNfbYYcd/NrZVRkkjgkjitRZzV1//fXV3k7Hl0CAwLqXXXZZYYCk7SN5i6dAqC+88MLEYna5igsyhz0LA/lbJnEtR9rjNfikL0EQULVaJP6EFTZqFwZ3+psZORcOh8RhOxdtIml/+gETiErvIUCw5UjeUGdir5hXUH9BvPhDQt+TQv/B9g17KnOcSpDwZAskzkIRFVSsSG2vueaa7GH63csIQJown2Ecpg0s+46TN0hXLIwX9Buk8kjwsZ99/PHHk88++6wqgs/1LOSIEzG0MdyXeee6667zeSc9N0HiGKNYfEDwIY0XXXSRxovYKPpsegR6ROAwIkfSweCNCBo1R/oFiU/PQArxstAhrt6I22v95MXnevxx7bwCicMWJk4sGMSq1AcBpG+oP8AWqRrqDVbI2QLhxqj9uOOOS8xT2FfI2WOq+c0gbwGAnbghbcmzm4TQIRWOKv2DDjrIB+5q7qNjq0Pg6KOPLrxnOCQ1U0FqA9mPfRXpjErfIkB2BEwraAOk5ry7afIWa8N4YaFAXIoPIa90kRfPz36ycLPIBU7mSi3iIHFI+KOdNVlBIIAqQqAVEOhRGBGSTJv9gL2XwUM54BpunmTBJk/fFv+Z2LxLeIC4r5ZPW8l1ez1ybOI2TqGO5sVWy610TgkErHP7Hhv43LXfpF5djmQfrvr81aOYyizYRFD2UvQ9U895WABbVReFEih7onbWjMBiiy3mIYDoE+Q1vvTSSz22X80XrNOJRgiCOUZ4MHAuSYgK4oip9C0CplL3GH/c1SRqHt/R1Kke+iNdE8YLs3dLb+rRd5OoeXzRchchywOhbEwQ4GGKCEFk9trlTtE+IdA0CPwvoFoNVSLFEZGxTSzucZhMfeVxuWzlVMPV6ncKAX3N/slj0XFVU7d6XLL63aGzrwSRghATq81WsMFsVIKpqD1WWyORYcIm9Rp/BAAl/hiBghXUuXdbxcI7eNw37sK7b/ZNgWDMjSzUwyQ/waSDXg0WcWYsr9iQDWiU/v37BwvXEUxV6e+laUMC2ROI7djIQh898cQTA5leTHMUTBLn2YAgkipCoBUQGNlCOwyutaJIwkh/ReBegrOa2NujYfOyMslXWgjkaV5EwWzbAisi8wbyCcBUZL5qZiVVaYFQIAXkehQCuzJgmOFspZfQcd0gwABH+5JNg9UrwXYJnknwXoIvm1FwN1f4724kNgziBOTlGrRdzKrANZCgVXotBmDa3NT6PjnQ3gSOXXfddbuV1lZUWR1UEgHaCBIHiWcMoA3JiECQVMaISgvnEfwXAk7+UzI5ELzb1G1OwquZWMmPafZPHgAcDYDFifMgv5KuVNoa9TuOPmD2qP6O844yV7DAYhGI5L7Sd5wFGoGcCRRNlhY+IV/0Gwg6f5Vei3OOOOKIYI4uvtAnMLQ5TAS0SvQXFSHQCghUPrqWeBpeTKKsW7wvT2PC4GuGoMG8A4vy1eWdzst38cUXBzMw9RyHTOZMxBQmb14kM34Pq666ahg4cGAgkTHbyxXzZnICyDHUjYwMZqdX7hTtqwEB0tBYXCYnTUy0DKZgjwoEyVe5Qhsz0ZstnefDZZIm2rrZqbj6nWjrrNZpe6K0037dEQHIpDkshDvuuMP7zdJLLx3IAEJEdpXeRwBp7JlnnhlWX311z1uJOQULJwu83e3NzanFzzUHA+8PUT0fT2QcIK2eOcF4nzM7qrgr95OF32mnneZ9kgmd45G0ZE07ck/Wxl5BgDHBQv04IYfo856y+DeHloJ6tdSNGSvuvfdeJ/WkaeQ3pJ5xhL7BYo3xCFU+CzbauzsSRkYG1P0QObI4kIMXQimCX6oVtL0pEbDBsscFRwIjWQVDUDyJcDAoVezFS4YNG+YxgTA2N1JWMII2kIq+s89WVh6CwNJldeshRIDXeA2TvvXYELbUM2j7f729zj///ELIFjxNzf6pLDQYKtsqPLHVrzuZEKfJJtlCm8W2MwlfYgOyhxHBCBpj43KFsBVEaud8YsXhdVbKcLncdbSvZwgsuOCChbbEQ7RcYRwgeG+MARbbvtQn/YRsH4QqKmfgjlc6HupcxyZyjxNYrh7a1zcI8D4yhuOJStuQeQMng1KF/mHEzbMsEEaIsaLUXMEcwXtPZgW8ortzRCA8CdejHnPNNZfCh5RqBG1vagR6LIGzF8DF2ayMkKCxGjKvxJIqS1bHBx54oOcoRHVm6HCJMJ3lo7MXySU4rJqRzqCeI2dqFJ1bCq7AfcyrsSj5uV/g//8hCTRPM1fH4cRgHk+e1Dp9jL7XBwFU26x8cRowwu6r3tieeXdAbYLtC2pOVr/0BSRr2KhhR8mqGWNiJHE2CbsNE23Iap12tSjtwSb7XDUJ6hOugYSWT1S83Unt8uqobbUjgJSEfLgUHAaQrpQqtL2RLM9jnD4GySuSECS5SFhIUI5NK/ZK9C3GBVukua0t+Y/zpGr0Be59ySWX+DhgsebcPIPk9iqNQ8C8yL09mSdoIyRwSL/yChI2ixXqYwXtz9hB+/P+I2FDMo+0jHHn7bffdlUqeVORziHRp7/YAsIdrPKuj0rdQhq5tBhNkMWv9GtWo6bPu662CYE+RcBeih4XsiAQoNcqnpgIu2TYBiR1hJ+Ix3I8OVFZMdlL5IEaWaXxZ6JtD856+eWXexgAVtIcby+tp1QptcJiZb7ddtv5sRw/yHLkqfQOAsTzIwYXbUMMJcKJEEwzr9gA7FIxAq7SLjZQejw3Vsxk4yDUCHHdcOMn3y3x4wi+ycqa44nVxDHpuG/p+yChI1AnQZ6pjy0SSvbD9Hn6Xj8ENt1008J7hxS+VGEc4L2kXeMf/YIYbXnvNUG4iTuYlu4xDhAXkGvlFaRwxIrk+hxbrj5552tb/REgBFAM7cQncd7yJKm0KfHZCLzMOEEbmiOEB+1Gsk5oIsIWETCcGHME/CZFF2MJbc3xSHX33Xff3PiAPBlzDFLaKA0kNmE5rVH90dAVhUDPEWBVU3Mhlg8icFsx+6RpEg/PsGCr69xrEsQV1VgctMltR9wf1GqlCvcgmCORu+PLzEtKoMdS55ESJR5LMFd+6+UshXBt28H+HktJQ25L2hMVB4NrXoFcEfUcEsaxMQo/kfIJvpudhPkNUSMIK5M0gXs5DxWJSWGd7OfdxwybnURyLKm7SKfEvfMmibzzta02BHhHiaofVaGMAyQIL1VIqUQbxT9SKbFg667wDkfVKOeyaID05xXa/P777/dxANUr6Ziok8aBPLR6dxttAQlHbRkXWHxnXM8r5E3GHCKO4WbP6gSeRT7EK1sYL4gRSWBn5pRIyuiPxCot1bdQz6KSpy+ZVM8DCZsELzeWafae+i0EmgGBqgkc9khIQoi8P8CC95rqs5C3zjx5Sto0kC6FST7aO3E+L0ulhRcX6V18qSFmpk7NfdmYUEiUzovJ/UxMn8wyyyzJcsst5xOAOU7kBoGttC6deBz4k1OWVTTRyiHQRDjHhpE2IXp6nnSMwZWJ09RX3h6QN/NO9Ak/S9yyuDLwQwz222+/wkBLHyMocN5ATi5NUm8htWNyJ0AngXyxmSK9GvXozpYuWwf9zkfAnJWSwYMHe3YNsmPwjkVChlSddzCvmLqrYDPJ8aRCq6ak7dt4t5HSsgjMK7R1HAe4Fws/JmykPywM6BOlFpt519O2yhEg2DfZUxgnsIUkCwsLOGzYWJBhs5rXR5gTNt9884KWhnGDMQcJfneF8YL+hTQtkjgWmBC7vLEGaS9zAv2CRYd50bstN+dDMMneAzFUaTwCzAMQbv7gIKaGr3ulkAjTb1upVEzgeAGOOuool4IwOdLpeRkjIWOALOe8YF5gPqlyHC8KqymMVKsprOJYmaEi476obvMmcq6Jao97xT+O5zzIBi836h4N3t2jTxuRQQPSRA5bHEPIP0u6qkim+Q7eDKDZwiobhwVIFcdD+lF5VFPIr4o6DPJHv8P4GTKZLdQVNRzknnanf9LWSO7Iw7jSSiu50Txq3ry6Zq+n310RYBzgvYttH9+v9Ceqq1KFHLnxWMgVhKzaQp8ibybXoV/R9/IK7zeq9Hi/7Cd9A3OLUpL8vGtqW3kEeK+YYCFhLNhJrQa5p514H2kDFntMyHmF9FdxoY/UFNV5KQla3vlsQ73KPBGJGf2DxV1eIV9ydGZgjqBfM86YJ7wv+hEaYCqi8SIPvb7bhsAn/f6Sb9nsouu6IGdxx4KDgnaIHLzNXiomcLBTVrtpEOnwvJyscpjcSSqet6pikMamKb7AqFJrZdBM0OYo4fWASKCCyyu8dEhull9+eV/58VKm687L3Z3HZN51O20bA9+1117rKsk0frQltowMsqSuIj9ttkCo6BPR7oWVN7YqeX0ke276N4Mn6lTII3VAmoqNTN6gil2MhZzxPglxQ0JIP411x5aGc0upb9L31feuCFh4kML7FzEFX1RjSOIYaEvZQTIRRzU659Kvain0H3NQ8HalH3LfPAkN/Q+pMCSOROXcOzsOsPggP7NKfRBAGoq5Sxpn+geLfsicBX93M5s8QgWRRjIaCRXSWRb61RbGhXMtZRp9kn6GiQ+mPvSHbGFxiAqfdGtokxjT0uMFRIHxQpL7LHJ9+5txxbJkeDtYLMCE1Ji0LVEn6lXSBA5VfKXagbx5qF516u46FRM4VqqQHkBjQkaihmgbI1LE3gzOpR4EVWckf0z4NEDey9RdZdnPi48Ej4EbsTci1TzxONdHOsfATjJ0i1PmA7XFIipM5rywKuURwGAY5wTanUGYyRIMcTxAzYADCm2Qt0pGAkIyeeweaSuMjM07rPwNS+yFkJMEG0LGH5IcJLLZwuT+xBNP+CRhscBchYpKhJAkkfhD6u8x+z2V6hHA2SAuxMyTz9Xp2JhClPLew/QdWATGcwk5k0e60seX+05/wG6JfsmET5t3V+ij9D/GrdinqQ/9QaU+CCCpjySdCREziyFDhng/QW3KeAIZyhv/UX+iMqVNINbY1NaqJaE/IuFF68J4YZ7suUIDxgvmo+eee841AwgIWGSyKGW8o39BAEtJDOuDmq7SHQIQOAQ26YJTE4v6epU0gav0mqjhWXQ0qlQcRgR3fuvs1p+DRzUnyrlJ3yqKlk8oECNTfi7R+smNZ6sc/13tP1vZBbOxCUYcvT64fxPA0V7UoktxfXsBfRuhTWw15lG2zRYvmGerbyeqt0p5BMCIqPgUMNxoo42CDW4eXJdQHaYaKRlc2SZMd+k30u2hRkxS62FHyt8xf29sP8IIUCf+TIrWJQyBEcVg3mv+ZwsKD0diE4ZnCSAEBXUitAz9WaV6BN6yoLs2+fqJJtXyYMmVXoUo/PFcm1xLhhqq5HomKfE2pj6MS4wDtHu5YirTQnBoMj6YBNDrQ3gLlfogYCQs8Ech8LqZqoRFFlnEw72UGys4nnfShAHeJgTxNmm5jy/sq7YwTsw555weioq557XXXvN5woQQRZdivCAkCX/0TfqSkUYPUUJmIL4Tusq0SEXn6UdjESCbB+0CD6GsssoqHl6IbYSoIpEAgb8JQXPooYf6PMSYY1EOCoGjCTdjNtJ+jAmWPFNIfCpTu3uf3X///X0TY5fZ/HrgehIEcB7XI7QVoZNMbe+5dE1TFC/RJ58VEzhjmEUV4kXIkqaiA1I/4kvJJpPE1Uze4iV52WyV5j/T14778z45nr84geQdo21dEWBAg4BRbCXrgzLpZiopphLxDAv0HSZPXpLsAFrJdTiGvkaGBgZmU6t4fC+IWbnCORB+/rg3EwiFAd2kReVO1b4KEKgWQ97VWGy1G7/W9Mm7HDN+0L/Mu7Sq68TFaFUn6eBuEaBPxH7BApqFFwv2Sgr9gzGDMuWUU/aI4NM/eOchaBST2HqeXv9R4h/nMEbwxwKB35T4PCVO0+Y+QgDhi5li+SKc+KD0EWL9UUyj4mSbOYL+htCBBR3CHbO5dHJH3D+z0/fMT6R+NG2CZ+ohHiWLffNaLzwJxI/4pJS3bUHCsYw3ZtvphM/sPJ3HMK/RV1jMxr5WuEgffKmYwKXrkiVz6X153zk+noNkrFbpW7x2+nwRsohK73/ShtXgzUvBCpZC50YS2pPCwGpGxn4JgoFWMwlT73Qf7Ek9dO5/EUi/h5VgEidEjq2mH5W6Nv0rlvS14zZ9NhaB9LhfSU0gSvEdZaHX0zZlvIh9NH3tSupSbd0ruaaO6RkCtKeZ8LgEzZxLXBNoqvbCRSFc5vTiv2k/M7PxdJr//Oc/XQBAgHAz/3ICSNo/JO9o8hAKUEgQkFcsW4z3RUgdhC1dkNJCLM1xK725z77XROCQbMQXo5KakssySutIMt3TwTt9Ddh2NS96JACV1FvHFCNAm1fT7rR5VGMzgPZUDcHqPK7QacdqpHn0kWrqXvzk+lUKAd7lSvsFq+NYkKL2pHBfBk4KbZseyMtdl/Poi+oL5VCqzz7apZqxGUlGlGIwubJI60khP3Ik+Zh7VFOX9BwXSWVP6qJze44ARIucxqWKRRko7OL9RkVqdrZhrbXWKmw3J5pChidU+5G8FQ7I+YJ5hoWb6ULecg7t800VEziYJ6DQmbEngLmaV0hF9mwcB3umwGJRyWWZbKVPDgngGryYvJDmkdjti4nKDJEpdlNpMSm6bJXyCIAxZClKvBgUUVcx2HZHhiFvYEy/od3eNlE0k2ccpMvfuXgv7Y0ahPtzPQZkbDDLFc7hvqhaIQxRGki9aqlDuXt1yj5IGG3AOGCG354aDXVoXKQx8ZXqF6jewZ0+YEGXfQXcXRuWwtUcWHwcYD/3s9ATuYdST0g/98T+kcUfdnOk16JQH1QxKvVBgLbgj/GCNmISnc7SJNLOpfpFvDP9KErpOY9xBsJdDfGK16LNo50051divsE5jBWMM+b0ViCQsW/Ha+uzORHAxCddGG+Y781pqUvfsxh/wUKFOI+IwqX0uenvFs0gWFD43LmLvkUfbVSpmMBtsskm4V7LdchLaQEw/eFhr7wYEDQcCyyNlq+EmWDThf0YpTJ4MolbpG0/tpZJlAaBQAIaRq4W2LUg3UvfkxWceSG6gSH35N4Qx/Sqzrwo06foew4CdF7z9PE2w0aFtscpBXuAaJPGAG2eZ11IESSddke0zeTJgIqhMv2h2sK9OR/HBQgY/S5P6sJEjYE6k7R5ovlgDHnHCDXaYFlkd8+nWG0ddHwIGPViM2LeV55r1LyMfRWLbStG66gpBg0a5IQuixeLLf5oC/oBgyJOMdUWVOcWzsRJOQMo96X/ZQvkDeJ+8skne59gHKD/sCqPhT6MwbNKfRAg36yF5AgW6NlJOjZKzBOQZMZrJlULORUga1lixljB2ED/oK0g2RxfqQ1d+gksaLcbuTPeYzNNnu08iT1zgoWTCubF7GMGxvHMHfRvxiwK6rpob5m+h743NwLkZMdpAe5iUTScJ5iXsTs+MEYdf/zx7oxAf2SOYu7IKzhKnHfeecHij7oNHeMKY0jkO+Tdpb+xsGUcgpP0WbHKVFRwl11vvfU8IKNVzt2rjagVguPapJrY4F4yXg55Kk0K5+cRioAwANUWk6Z4MFZjzH4dQlmYdC33MgQd5n5GEguhC2K9+SQyeE/CGOTetA03WqdMhg0b5jGcIn62kvZo+jbYuhs3btR5YRyMZHscOGIsca4NpAlplGwCrgopk6R5YEVCgXAdAn0Sm8leuC7XIYgnQZqJTZgXB47QE8QQ47lUqkfAJrXESJfHyyr1bh188MElLzzYsjcwbtCOhKSpth3oU0bKPJQQ12DcYWzJK4wNe++9t9+LY/mLYxZjgxEL32+TfN7p2lYDAswTRpgLYaPA3Iiah6AinAchhYjdZervLlenbRkfjNx5WxG+gxA1vP/VFJO0e7Bv+gb3t8m6ZD8jnpgRS+9HRvAKfTP2FzI1MKZUO2ZVU18d2z0CeWFE0mfR1kay0pv8O2FhjLx7u9K+ZOcxYY7vI5QUoUhiP6HNYyBfjmMeiYVrG4n3/kQ4qhisnPHLPFGdZ9Bvqw1SH69f6yeqkIoLsXWMtSbmsuvBVI21FnV4UugwuOYVE4l70Nc4gBIkrxT5yjufbUTUNqmOg2gi+cQ8UXJfbiZ202/7cTQKDWcrQA8iaWzc8+PlxRArdd9O3s7gSZsS8w0ixmDHIAymcZAzKYZnW8gbaAmYy7m8JBA/8hqaFLUqSAm2WWkmBtIjUR/qBtHnZePFs5AC/qIxQfAC59W1qkp18MEmofD0ZATvZpKFKINzJGbEaWQSzSv0pThB00bE6GNhVmkxKaoHC+Zc7geZLzXmcCx14VhIhElynNiTkYNgs8OHD1cWhkqBr/A4SBiTmoVqShhrIekQ5fR4wYRKTECOzRbmCYK+szjgHIJyv/XWW9nDyv4+4YQTCnl5maMIPp1H0pkn6MNRsMD9qCvzGHECjz766ISxJ2+hWLYC2tl0CJiktS5jPrmU8+aORgmDqiJwtAoDM6twJGgEP4ThMogyQPIiENXcbAm6NCAvKwESI9tlMicCdiUrcMCBvKWje5MCpdTAT/T/eB9eSPI2MsizOuScvJe5S4W1oYAAbUcwZIJhkkWDgQ0CvsIKK3iQTNqS9Dh5UlUGP3KQQtyYSGlDBk1Th3Y7MHIuATRZ7cRJ31Qqyfnnn59L/s1+JaFf0A9Z7ZvxaoI06IwzzvC8l6zmOUalfggwFjDBmjrUF0m0MZOvxdoreRPzCitKxQU5Z3FY7r1kTDG1WoG8cR8WEgSSziv0HSZuJmWIHtk7CCIL+VTpXQQYL1icM7Yz7jIeE5Q35k5mbGYcyZv0aLcLLrigEKSZBTtR8Rkv8ibO9JNwT8gbY0VcTEDUS80xXJOgwxwLuacvmX2158Nk4cl8kUcy0/fUdyHQSASqJnDpyvJCIVomWTErcAZVcuCx8s0rvGAM1nE1xsSPVOeKK65IYLbZAmsmJxkr5ih54x4QgFL34BoQQ47jj1Ugg4JexCy6tf2OgzNkDQkGki1wJsk8yZ/zCu2IypNI2hwLiUPyAtlHGpZVT9BebL/pppu8v0TyhqSH5PTmjJJ3G1fjIvrmHqzymcAZhLsb+HMvpo1VI8C7Dfb8obYqhTvbMX/g/Y/HI8lj4iaZNKSQsYI/Jnmzl3M1PtKceLw50STkqSx1D85DBcLx3IcxSqXvEWC8oI1Y+GO6QE5i2oRsLqS9yyu8syy8WHxzLGM/8wSmHKheGR/ShUU5+VfJFsQ5kbyhKeIepcZ+VO9I6LgHC1CuXao/pe+n70KgWRDoEYGLD8FqGxUlL4J57CT3mG45OynHY1kBp0kcai5eUCZcVmhIW0iTBBEkYTUDdbR54/obbLCBD/KlXkpE3qRh4VgkAaRvUekdBBjwkMQhfWUgpF1LqaaR4LE6RhJC20DiUHNjD2MxeTwvJvlOWakjoWE7+6PUFTUoxDwviT1Px6SPhJeUbdSHY/PsbHoHCV0VBLD/iOoo2g3JS6nChL7zzju7pDROuPQLc4hJVl999cQcJXxCZmI1hxXvM+znWCRvpFWD4OUVJmHU9HFyZjxA+qrSWASwJTODcG9LJF6YO5TSoiABI4dtJHEs+plbmBO4xgEHHODSdt5z8tz2N9OO9CJ/1VVXdQ1RluxFBJDmci3mFuYJpPql6hLP0acQaDYE6kLgeEnMQ6OgtuQFKych40XB0BWj9jh4M+nykjIB8Md3tjFo82cRvf2FzZPUpUFlVR7PYdVW6gVOn6PvtSHAJIwULrYj0jikZqUKUjvUmTgjROkL6hRIFxIYCBv2J/xGBUo70g9wSEDyBnkrtUJGPYaTDQMyiwmkgdRPpe8QYFFFXtH4/qH66q5A3s370N/59PserxE/2UefQNKLqqvUAo77Mb5EaSD9ASPjUv2mu/ppf/0QYFGPFA0CzriPJqWUFI67stjneCSvjAP0BdqTfsACAYKO5od9cR6B5DEHkPO23Nh/zDHHJNhRc03GHgidihBoNQQqDiNiHb1kMZG3u9DaIOnH2ABaNjaKvYDBPBc9ON5FF10UjAR4LjF7wYvuYYO2u5WT58wkcsHE74XAsEUHpn7YKqzwi1AFxBMizIBK/RGwgdNDediA7EFVzdbE4yfRXnmFkA24dBMCBNd90qGYjYyfS4y+dLGB2d2xSYditnYeddtUqelDir4bOQyEAKAPGin0eFL0H5W+Q4Bcs/xRbELNDe+TrY0RrUAATpOAeA5DcgoyDhhB80NpQ1vQeV8w9bi78tN/yhXOoa/ZZO/XIewQORLJw6zSOARoV+YKI1YIDjysU3bMT9eOccW0MsE8BT2XMTkuCTFi0vb0YR6OhDAlSy65ZLBFQzDJWrfxSekb1IFCaBmT1ntoE5PGFV1bP4RAMyPQD8bZ0wqaOiOQmoIXy0TegdhQ5CCr9GWA8DFwm8G6B9xl8CbGFzGjiANkq62Kq2irtrDMMst43CdOMuN5D+THJKBSXwSIrzZ06NBAqhHiKZE82uxKgqkvyt6IAZxJlXhtJlVzEkcATTNi91hNBP1kkjbJm8cXJD4UpKxc4XrmAh4uvvhin7QZyC2EhF8HMqHSuwjQdmb35sScO5kqNNx5551VxfyL4wBtyWKAdmM8iXHe0ouz7p4GIkky68cee8yJnEkGPc0OpEClMQgQ94/Ym+SRpB14X00K1+34zqKMvkFgXxbk9A/GHsifSeH8HSeQM7HauK5J97t9QPoH8b3M/MfHCxaJpFwiZpyKEGgZBCBwPSnYMaDetAd2MTZhGvK8i3pyj2rPxd4lOlVQr0MPPbTaS+j4bhDA0NikaO7FBcY4GBx++OFVeXmi1kJNghqFOHIWuNNDw9gA7dvLqUCy1bPB3J0WjLx7P6Q+Fsi5pE1e9nz9rh0B1JmoxukH/GECwbjQyELfsYDh7l0Y60QMOvqcSt8jgPkDDmVR3YlTCR6q5VTh2VpyLJ7K2LviGIVHOeMQphLVtivH32O22v3Ndg71K32WSAfyUs+irt/NjECPbOAwTMd+wFQWPnAT4LOUy3Zfg2ARmAsTCsatuIyr1AcBJkdi8EV7JxxNsIEkhlMjC4MvRAIDaSZtU78mlmy4bHiKRta3Xe6Ns0iMvcdYgGNLMxRCj2DrhN0U/cHSupV0gmmG+rZrHVjQDxkypOBcRpy3UjE8+xID+gfzhEntvH/gMIFDHotBFSHQCgj0yEiIfHGIolF5kgLFJDA1pT2xwbXuZZ999nGbCC6M+J1UFyr1QcC8/zxFGeoH7OCwLcKWybIc1OcGNV4FdQrqWxIWozInZy4pcmwCqfGKOq0SBMzb1FVaHIu684gjjih5GqpWk5j4mFHyoAp3MO7QF7meDbZdzkKVhq0tNlQU7K9Q26v0LQKktMMG0RzbfH4gvRb2zNgqNrLE/oGZDd9R25Oj14hdI6ulewuBihHokcUmdknRvggbKOyZsF3K2psx0JpKxV9i8xT0vHQ9eXmxlzPvJL+PhSJwg+VYj/jk5Fs11Vz82W3i88KB+tItAkzC2KMwedIHsFfMy0XJhZg0TfrpEzY54rBvq7VAxKP9C2QRO6usnSX2ktjBYKTMQIxNZDlD6VrrovP+hwCY8z7znoO3mTC40wltEAsECxtZS0kT3rY8lxYOyMcBJs70cfH4cp+m/vK+B0HH5pH+B1EzCVvRtSB2Fnzac1xyPe5j5h7lLq19vYBAHJvpA7yLEDnGkLx54l7Lt808Auk279OAw1utBSc25h3uT85dHB1iXeI1mSNwYKBPUT/6Y/aYeKw+hUDTIWCdtuaCaJwgu9g12IO56grbF2wSYiHo6lZbbVWI0YO4mvACiKqxZ6i0YLNgqyOPDxZjA3FPUuWYsXTRZbBtiHZ5HEN4C2ytVOqDAKoHAizHiOe0x9Zbb51guxYLx6AmIZwDAXzJ1kGYj3PPPdcDs9pkHw8t+8lxxIQi6C85V2lvwozYqtnjxaWzPxD8lywRxP2yQdhTaBFXrlS8sLI31s6KEcAOidAwqE9533j36B/RhpH3nKwtxHRjP39xHCBLCqpv2qhcn2Af4wrhiQjQTIqlqPriviZ19ZiPcezhk7GIsBLcjzFq2WWX1ThQcavW70Bs1o477jiP6UZb2cLL81ZixxYL5jgE1iVMDOE9yFnM2EFbp4+Lx5f6ZJ4wT1UfB4jzRvvTTwYMGODzBONSLMwnhJqyRaj3EXI7E7sw9tt4nD6FQLMi0CMbOB6KVEckpycYIgMl9kfmheqDLYFZietlK1/fFwdvPjkeGyWCNTLxlhq8CQxLYnpeNAxN09eI33FYwMYiGi6nyRuDhal6mxX/lq0XbUa70N6QJTCHxOGQAFm28DAJDgVxkqWtiN/EwIwNjKnZPLhqnHCzQDDQkoKLPKprrrmmE0Cuxb24lq2UEwZcAv6S/sZCiPigzfU5hnoxIVBPld5HACN1ouynSRzvP5M3WTgg3fF9TY8HcRwgeHeptmJsMKmdJ57H5jaONVyPa8U+QR/E5g07XMagSN44njGKiV2l7xGg/bCPxTaS2G2RxBE8F3KG0xmZGVgIxv5Dm2K7TABviDeOaNgxlyJXLCIuu+wyvw7psRgr4rXieGHaIZ8nmFMieYtzCvchf6tJ+fseIN1RCNSIQI8JHPflxUqTOF5EJt20JyiT9wILLFA0+MYB3cI9lJSSMAnHlywez8sNoSNgaxy82WdxgArG1PyGvDG5q9QfAVa6TKoQsUjikHxZ6AaPsE9wVkgWgygratoithUTKkbvpDoiqXVewSuVgNAcF9uf8/EuhaTRn/jNvZECMwFwT+7HNjxiqR/1VOkbBMzOqYjE0Xb0h9jutA1jAxMygVjpH3Ef7yskLq8gnSPzQvr951z6FUF6afdI6iB0BISmL3J8JG8QTJXGIQDxQkJP9g3eZ/oC7cYYgIYkSsHYh7Se/hHbG+kpv3fccUf3XM17Csgb0vl4Hc6lb0H4mSe4H78hdmT6YHscVyBveNRXI+nLq4O2CYG+RqAuBI5KZ0lcfPn4JH3S1Vdf7UnuSXFDzjpW5HHwRlJTKvQIE0AcnJkQcEVHrI6nEGmzUMvE/el7SvLW+10pj8RBrJhYaQsGU0i12T25egs3/cUXX7yQ4ojE9Ay8eYX8uDFvItck1RqDPypRs2vyVFkMvPE+rNYZpPGIhVSKvOWh2vvbsiQuvpMQLtScvP/0Gz6ZkJGixmMgdnmFsYExguNoY8YTwlCQ/5JrkYWDvJfpyZtjI3mT5C0P1b7fFknccsstVyBxtBltSnsxJyBBhWwjvaV/YH4RTXQWXXRRV8Xn1ZzUWlHzwifzAhk7GAcw1+F3JGyRQHJPxhDGIJG3PFS1rdkRqBuB40EhcdgmxReOF5PQIuTAy0pCkLxEFQcrMWwgsoWVN2pWrsM1sXXLitBRwWH3xMTNC8kfKzGpTbNo9s7vSOKQlCIdi22ABASijhoU9QbtRhujLiHfLW0KISduYLZN+U0MN8I+cBzS3Ntuu81zm9InsKl66623ktNPPz0xY+cCgUfyhhpX5K132rrSq0LikMhDqpmgUZuhPssLMcQ4ALmj3zAOoHLNFuwc2ccxtDHqr2zBjo44b/RB7ss9CXMj8pZFqrG/ebdpc3KVsjCjTXnHMasgDR+pEhlTMKFA1UksP8xwWOyjAoWMZc1tOBabSNqdeeKss87yfNmoQzmW8YJQNywg6R9RcICqlYWiyFtj+4TuXjsCPfJCtZevqOBlaBISj8BOeBFcxUlvYiueouP4QbgBm8Dda400SLia28qpyLXcJvxgdjHu3YaHGZkZTEVSdC2ubaqXYJN8sIHBPRBtte4RuosO1I9eQcAG32Ar57DFFlt4+9AGNmh6BgXaHs8vk4T4vU195u1DextBcw9VvuNdiidjLPymP9jA6l6ueKOR2cNIejwk9O/f39MqmWTOU3LZYO+esKY+8/pQL5XGIMB7b4Q7mIOCeyvbgioYocutDO804wCegLQ57zshiWKhL5EijX5iE6/3p/T+eBwhZA4++GD3NsSL0aTA7smY7lfxWH02DgHGbyNk4dRTTw0WozHYYsvfZTOJ8QgGRua9cnzyx7tPCj3aFO9V+oERsqI5Bc9n+ogROQ9TggerEbXCXGKkzq9BxiAyLhD2hm2MFfQ/I5KNA0R3FgI9QKCuBI56MHASKqS7woBL2hLcuAkvwAtI6qR0YTBnEqdAErh2qULKLf5U+h4ByBIDpqkpgqlIfaKFbKUJV6wVAydknPZnMCYPIW2cnmiJ28Z2BmQmd65tq+t4Cf9kMqc/MAiTwstW9n4Mx7FPpbEI0Daksuqu0O7TWTgYCBzEnTAxtHkstCvpkxgjmPwZB9J9JR7HJ/2QlEr8qTQvAizoWIARSob25bdJanPfW9oaAsd+xgXGDMaFtFAAYscCziR3TvBN+l8gb2kUGHNYSJj9m9+La2ihl0ZI31sNgboTuEoBYJJlJc0qjBePuG6mgi1Ia7gOeQx5WSnEGctK33yH/jUFArQng6zZlJStD8eRqN7U5y5xIcCn2Uf6xBxPJLG9hZ1wSR5S2VIDMsdDCLmWSmsiQH9g4UVQaIpl0vD3Pj4NE7yFo/GfvP9I+TlHpbURgDhVIvniuEjgiClptnEeS86cEQoAIJ1DAEAxB6ciclc46P+/MF7wpyIE2gGBhhE4wDN7uWDhJvyFNLf/wF+pstRSS+nFKwVOi21n8GUFDmEnGCuELb2ihrQTKJiCxIXg0CrtiwCJ5gnIiwRu+PDh/pf3tEhQBgwYkLdL29oYAVSuSHSRypsddDAHmCLJGcGBkdBSkLCxkFQRAp2AwMhm+Du4UQ+KVA31iBkau7SlVD3WXnvtsP/+++eq5Eqdo+3NiwBkDbsViyFYUKFC2OIfKlUG5elMtWbGyR7VX4Ny87ZnT2uGJB7VGPZQkHcz6S26JCo21O70BVK2RZvKooP0o20RwISC/mFRB9xu1hzXPBOHOTT5JzZx9BnzUg3mjep9RVLatu0OerAUAv3wf0j97vOvTNSsug866CAnc+kKYABvARyDxfiqSNyePlffmxsBDNbvtbQ5FszT1ehIX2LBjg3p7Oabb+6fTN4akCM67fmJqvSuu+5y5xZyUqYLtrI77LBDQAov8pZGpnO+Y/+GU4yFGXGVOqQtFpyjIPY4Upm3ssaKCIw+2x6BhhO4tkdYD5iLAOsGJG6oUZG8MIHHgo0KKlacWrBvE3mLyOhTCHQmAowXkDgiE7DYS8sdsI3ETtZCzMjMpjO7R8c+tQhcxza9HlwINB8CSFZiKInmq51qJASEgBBoHgQa6sTQPDCoJkJACDQaAUs+H66//np3bGp0XXT/6hAgXh+SMaRk/OFUwCfbiQepIgSEQP0REIGrP6a64v+1dybwN1T//3+XisiDFvteSFGWEqKISkh8FUplCS1KyzfaRNlKKCSRQiFKuyLRwpeStKFoRdGCNkm75n9e5/e/t7kzc2fu/dw793Nn5nUej/rMnHPmzDnPGfe+7/u8FxIggTQIII7XlVdeKSqZeEoxJNMYml0dCMSEK7Oglepxsn5mG1bzLSdNmkQBzgyExySQRQIU4LIIk0ORAAmkRwAaN5VyTXsZ4kqn4M/pjcjeXgRg8D9v3jyvbhm3IyOOSqGW8TgcgARIwJkA8w05c2FtnhNAlo6lS5fm+Sw5vWQEEEYGYUE6d+4cF97QN1mWhWTjsD59AnfccYfvsdLwHGfNmkUHpPQfD68ggZQJUIBLGRU75hMBBHU966yzdGDXN954I5+mxrl4EEA0feRLnTNnjq0nBTgbkqxXwMP72muvzfq45gEnTJigPcnNdTwmARLILgEKcFngCUNdc1yiLAzJITwIIB0XvohWrFih868iD+v777/vcRWbC5MAAjT369dP2rdvH099ZJ0PBTgrEX/OERgd8dP8KHi+ffv29WNojkkCJGAiEGobOBjQIk8eYoyZ/0PwYPN57NipPpU6xCSCNgiaBZbcEUAOTcSQQ1m0aJEsXrxYB30ePny41K5dW9fzf/lBACmQ8KX+5Zdfuk6IApwrnqw1QoONJDxwHslmQSw25LVmIQES8J9AqOPAIVF6kyZNtHu73yj/97//0dvKb8iW8QcNGiR33323pVZ0tP7evXvLsGHDdC5VWwdW5IwAPB4HDx4s06ZNSwi+mmwCyMwB2zgW/wngh+txxx2nU9pl627z58+X888/P1vDcRwSIAEXAqHeQq1Xr55Mnz7dZfnZaUKKH8Y6yg7LdEaBBs6p4IsJWoBatWrJddddp/OtOvVjnb8EsL19/PHHy9SpU1MS3jAbeqH6+0zMoyMt2bhx48xVGR137dqVwltGBHkxCaRHINQCHFBceOGFWd8msCJGAmWW3BNAqi23gsToEydO1PkRhw4dKrt373brzrYsEUCScRjJI4zE5s2b0xqVW6hp4cq4M2xHW7dunfE4yFcMQZ2FBEggdwRCL8AB5T333CNNmzb1heoJJ5wgbdu29WVsDupOIJkGznoVgo+OGjVKC3Jjx47NyZa6dQ5ROV+9erXUr19fYH9qzleZ6vopwKVKKnv94NCQab5h7HQcfvjh2ZsURyIBEvAkEAkBDrkVkabHD6+rIUOGeEJmB38IlC5dWipUqJDy4D/88IPceOONUrNmTbn//vsFDios2SEAbecNN9wgLVq0kE8//bTAg1KAKzC6tC9E9oQRI0boWHwFEbZjN4S96TnnnBM75V8SIIEcEYiEAAeWlStXFhjYFilSJGtosYWHQKQshUcgVS2ceYYIAgzvu6OPPlpmz56t8zWa23mcHoG1a9dKo0aNtD0VQupkUijAZUIvtWvxjB5++GFtI3rbbbfpvKWpXWnvVaVKFW2mYG9hDQmQgN8EIiPAAWSbNm1k5MiRWWOaja2HrE0mogN52cG5YdmyZYv06tVLe+I9/fTTbl3Z5kAAsQ9vvfVWOfnkk2Xjxo0OPdKvohND+szSueLVV18VmH306dMnaSy+VMfDtuvMmTOlVKlSqV7CfiRAAlkkECkBDtxuuukm6dixY8YIEUT2ggsuyHgcDpAZgYJo4Kx3hPBx7rnnSuPGjWXZsmXWZp47EEDQZPAaPXq0jqno0KVAVdTAFQib50WbNm3Sn3v4EesW8Hr//VP/ShgwYICcfvrpnvdmBxIgAX8IpP6v1Z/753xU/GrEttlRRx2V0b23b98uSAqdib1PRhPgxZpAJho4K8K3335bzjzzTO09CWN8FmcCX331lRYG1q9f79yhgLVFixaVAw88sIBX8zInArt27dLmAgjn8sILLzh10XWwE0bIHXhtp1JgRwqHIBYSIIHCIxDqQL5uWNetWyfNmjUThDzIpOAXa7du3QTbqfiQZMktgR07dkj58uWzflM4SCA4MwKdstgJ7NmzR5AUHV/4v//+u71DAWqOOOIIxuwrADenS/BM4AmMZ4QUZm4F2ue77rpL/6iFfVyDBg1kw4YNSS/BZ97KlSv11nnSTmwgARLwn4DyPopsUYa8hiKclf+UZs9QW7PGm2++GVmehbVwlb4nK88w9i40b97cUCm6Cms5gbqvsiM0unfvnhX+NWrUCNTa83GySgAzHn30UUOZeHg+k5NOOslYtWqVbRlLly51vVZ5G9uuYQUJkEDuCSBWU6TLZZdd5vphFftST+evsjMxlLFwpLnmcvEQuNJ5Psn6Kg9lQ3nlGSqTQy6nH4p7vfHGG4ZKW5fRc1DazlCwKKxFKK2YAaEs2fsdq4dwN2/ePAPCXrLSrl07x3FUdhtDafeSXcZ6EiCBHBKInA2c+hBLKNhmgDF2qgVGu14BK1955RUd3RxbtG52J6nek/3cCWTDDk59qQlSPyHBdzZDzbjPPDyteNdhN6gEA6latWqBFkYHhgJhk88++0w74SCd31tvvZV0ECSwHzNmjHz88cfaAcsteC9SbFn/HcA+EblqYavIQgIkUPgEIi/A4cPoySef9BTK8KjwAYi+X3zxhYwfP94ziKzaTtXG3rApWbBgAeON+fS+Z+qJChtGeOYpTZ5PM4zGsBAI4JkNAaEgQhwFuPTeEwSmhuMBfsC4hcFBzlN4jELQQyDrVASwunXrSr9+/RImhJAxiPfHQgIkkCc9eh8yAAAno0lEQVQEcqjty+tbLVmyxFDGuY7bBupR6XrlqJCwBmwlqIj+RvXq1V2vi11fu3ZtQ8VNMlT8rIRxeJIZATy7GON0/pYoUcKYMWNGZjfn1TYCKmB2gZ7HeeedZxuLFXYCKuuFodIDGoceeqgnZ5Xr1FAhROyDpFDz7bffGkqo1vc48cQTDZW5JIWr2IUESCBXBCJvA2cGPXz48KQfiMWLFzd27txp7h4/xgeb2low6tSpk/R6s2ChtBPGfffdZygP2PgYPCg4AaURTYm7+RnguH379gW/Ka90JIB3Wm1H255HTBCwPgPzuQou6zgmK/8loHYADBUCycbXzBHHDRs2NJQpx78XFvBI5RA2ihUrZnz44YcFHIGXkQAJ+EWAApyJLIx68aVu/TDE+TXXXGPq6Xy4b98+Q+Vc1R+eTmNY68qVK2co931Dufk7D8jalAjguakI/o7Pzcrceg6NKEv2CKiwFY7P4eWXXzaef/55Q6Uvc2zHc7n66quzN5GQjbRmzRpD5ZlNyi72XleqVMmYNWuWgc+ibBSVL9WYM2dONobiGCRAAlkmQAHOAlTZldi2RFWQS2Pbtm2Wnu6nixcvNlL1jsRWCLwfv//+e/dB2ZqUALZ4Yl9iyf46bZFjG1XZbCUdlw2pEzBvuZmfAbbxYgXa6nvvvddwCv1yyy23xLrx7/8ngHA2yq7QQJgiM1PrMd5jlZje2Lt3L9mRAAlEhAAFOIcH/c477+htg9iHpDLmdeiVWtXy5cuNM844w/XDN3YfaJEGDx5sqGTrqQ3OXnECPXv2dGUMIeLBBx907KNyQ9IuMU6y4Af9+/e38VUG9MZHH31kGxQ/lJQBvqE8G+PX3HnnnbZ+Ua3YvXu3oRwOEj6HYp8T5r/4UYLPJ35mRPVN4bqjTIACXJKn/9BDD+kvFsQGU95bSXqlXq3c+43OnTt7/pLGhzNsTpTXGIPJpo7XwJe/+YstdgyW0PjECuysYm3mv4MGDYp14d8CEFCZTRydgAYOHOg6mkpFp/9d4FnALjTqBRrKKVOmGGXKlHF8T83vrEr7Zqh0ZlFHxvWTQGQJUIBzefR9+/Y1evTo4dIj/aYPPvhAjwnB0Pxh7HQM7UXv3r0dNRjp3zncVzz33HM2nioUgu0L7pdffjFq1apl64stKkSgZykYARUf0cYUpgHfffddSgO+9tprhkpdllLfsHaCjWAqjlB4r1988cWwYuC6SIAEUiRAAc4FFDzqtqhUQX4UaPWw5QT7OifhzVyHbZKuXbsa7733nh9TCcWYn3zySQLHK664woABtlNRSesTtu5irFVO1aSexk7jsO7/CEDwiDE0/0WoCxZvAvh33bp1a0eGZp5wenrggQeYKcQbKXuQQCQIUIAr5McM5wh4uCJMifnDOtlxhw4dDKQtYkkkgPRXKkCpobJkGM8++2xio8PZ2LFjHXmbDe4dLmOVhQC2/Jy0RjVr1qRdoYWV9XT79u1aw+7kXGP+93/wwQcbQ4YMobe6FSDPSSDiBCjA5ckLgBhz8MJT2R4cBQvzBzqOW7VqZSxbtixPZp8f00AYiq+++iqlySD0SDLnErPNXEqDRbjT5MmTHd/XZ555JsJU3JeObfxhw4Z5/mjDtv7FF1+ctge8+93ZSgIkEBYC+2EhSiBgyRMCyvtMlDG3TJw4UZT9kOesVPJqUb/OdcoupDJiSZ2A8tyT+vXry65duxIuUo4POqekSq6eUM+TRAI//fSTKE2bqPA3CQ0tW7YU5X2dUMcT0an0VNxBGTp0qKiQK65IwPDuu+8W5SHt2o+NJEAC0SUQ+Vyo+fboS5UqpQUy5FtVNkSiAnO6ThHJqzt16qQFEZXCSFQAT9f+bPyXQIUKFQRfqNaiUqTpnJ7KBtLaxHMTgZEjR9qEN7UdqN9bUzceKgLKQUaQE1nZvboKbyrdnigTAC0AU3jjq0MCJOBKICyqxLCuA3kPYbh85JFHOm5VqYebUA/bI4RAYb7V1N+Iq666KoFhjCkcIVicCSD8h5MDDrymWf4lAK/zs846y/H9ir1n+AvbTWzdw6aQhQRIgARSIcAtVPXpGYQCzRo0bCremWzcuNFzylWqVBEV20z/4ldG0J79o9wBGjdsRW/YsMGGAdoQaDhZEgl06dJFlJ1bQqXKBiDKG1gqVqyYUB/Fkx07doiyc5MZM2a4asWV442oWHla6166dOkoouKaSYAECkiAAlwBwRXWZUoq11sso0ePFpUxwnMaZcuWFRXxXlRgYFEOEp79o9pBJeuWxo0bi3XbVGlGRAVLpVBiejFWrFghyonGVPN/h8OHD9dCi60hQhV4f2D6oHIcy549e1xX3q1bNxkzZozUqFHDtR8bSYAESMCJAAU4JyoBqXvppZcEgtzKlSs9Z4xf9/ilr0KWCIQSFjuBqVOnakHX2qJidIny+BXYd0W94AeEyjsr7777bgKKypUri8opKyocTkJ9VE7AZe7cuaI8yUWFB3FddrNmzbSQ17RpU9d+bCQBEiABNwL8RnKjk+dtbdu2FRW9Xgtwys7GdbbwGITRebVq1eT666+Xr7/+2rV/FBuVzZvjdumrr74qKm5cFJHY1vzII4/YhDd0uuOOOyIrvEEjCe2tysfrKrxB0/b444+LiuMoFN5srxYrSIAE0iRADVyawPK5O7Qi0MjBNgkaAbcC2xuVF1RuuOEGbuGYQCEkxvHHH28TcFXSdVm1apW2lTN1j9Th3r17BV6SVuEfGjl4Q0ctjA3s/fDvR6Vxc30PoP2+9dZbtQZcOX649mUjCZAACaRKgBq4VEkFoF+jRo3kqaeeEthzqQCgonKpJp218m6VadOm6S9kaA42bdqUtG+UGrC9PGfOHNt2qfIOFJUX19OuKcysxo0bZxPesF7YfEVJeIOQr4JGS7169VyFNwj96KfS5mmtN4W3MP/r4NpIoBAIKE0NS0gJbN682bjssst0iin1armGMkDU93PPPddQjhEhpZHesm666SZHXkrYTW+gkPRG2iendG94Z6JSlLeyoYRYQ2nUHN8N87+xzp07G8omMCpouE4SIIFCIMAt1EIQmnN9S2x5Iaq7iicn2AbzKrCnQ3aHFi1aeHUNbTs0bs2bN5e1a9fa1vjoo49qbZytIcQVvXr1ktmzZyesEBolaG5VjMKE+jCeLFiwQJRQL1u2bHFdHoLvQiN56qmnuvZjIwmQAAlkSoACXKYEA3Q9UnNNmjRJp+qCU4NXwZcQBLkzzzzTq2so27H11bBhQ1G5KxPWh3As77//fmRsBxGuBkb66gdmAgfEGcS2apjL6tWr9fYn/roVxF2EI8eFF14Yqe1kNyZsIwES8JcABTh/+ebl6D///LNMmTJF51vduXOn5xxhpI7wCGpbKHJfTtA6QftkLfAiRPgWNztD6zVBPYcgbw1Vc8QRR2jbLqR+C2OBpg0aN2je3ErJkiV1P8RaZMBsN1JsIwESyDqBQti25S3zhMCvv/5qKI2coWJ4edr0qBfPqFu3rqFiXRl///13nqwgN9O44IILHPkoz8LcTKAQ7/Lkk086rl39ACjEWfl36x9//NFQmkXHNGH4NxD7r0iRItq+VGVc8G8yHJkESIAEXAhQA5d1kTh4A6q8qdq+CdHjsW3oVWDzdOONN4rKeylR8KyDxhKJyK32TwjsixhxLVu29EIWyHa8F8cee6x8/vnnCfM/5phjdHaKMGkfYfOIQM4jRowQeJm6lXbt2umtY/WDxq0b20iABEjAVwIMI+Ir3mAMDiGsX79+8tFHH8m8efN0eAS3mSvvVlHerdp4feLEiaI0eW7dA98Gmzc4LlgFln/++UeHa1Fam8Cv0WkBkydPtglv6AeHGCsLp+uDUod8twgJgiwlbsIb4gMuXbpUFi9eLBTegvJ0OU8SCC8BauDC+2wLvDKlsZWFCxdqo2wEaPUqsIeCDdCVV14pYbWJAoNRo0bJ0KFDbThUKA1RW422+iBXwOGlZs2asnv37oRlwKEFKdzCUOCcgawkyKTgVipUqKCzmCDwNdOpuZFiGwmQQC4JUIDLJe0A3uvll1/W2R2WL1/uOXsIbxDirr32WilTpoxn/6B1gMbttNNO0+nLrHOfPn269O/f31od2POrrrpKO7qYF6DsvrT3LbRVQS7btm3TTjnQquLHSrKCvK7wtEW2hRIlSiTrxnoSIAESKBQCFOAKBXvwbor8jUjThe0jr4IvvksvvVR/+VWqVMmre6Da8eVfv359sW6bYs3Q6NSpUydQ63GaLGK7YbtQOaskNGPbHNk7glr27NkjY8aMkQkTJshvv/2WdBnQsiE7Cd73ihUrJu3HBhIgARIoTAIU4AqTfgDvjfhniHeFlF3QSLkV2NYhBAccHo466ii3roFqw9rPO+8825zh6LBmzZrAO3Z06NDBJqjDDvDTTz+VsmXL2tad7xX79u2Thx56SIYNGyZeYXPatGkj48eP104r+b4uzo8ESCDaBOjEEO3nn/bqIaQgNtbGjRu1cOZmzA4vxgcffFCOPvpoueiii3SO1rRvmIcXwObNabsUwi2E1SCXmJG+dQ2IAxhE4e3FF1/UGtPLL7/cVXiDZ+3zzz8vMBnAO85CAiRAAvlOgBq4fH9CeT6/L774QsaOHSszZ84UlSvSdbZIeN6pUyed3QHBgYNc4HmLtEnw3DUXrBHbzEhHFrQCTRWElw8++CBh6tWrV9frLFq0aEJ9Pp+sX79eb+EvW7bMdZqw1Rw+fLgWyN1+jLgOwkYSIAESKAQC1MAVAvQw3bJatWra2B0x0mDwfcghhyRdHgzGEbIBaZnatm3r6AyQ9OI8a4DN2/z5823bpVgj4uOpAK95NmPv6WCb0Sq84SrEBwyK8PbNN9/okDhIgeYmvBUrVkxnUEDcwyuuuCJUYVG8nzR7kAAJhIEANXBheIp5tIYffvhB51tFDDGrob/TNFu0aKE9AhEcNYgFBvH//e9/bVOHBg6aOGjkglAQrLhWrVq2bcaTTz5ZXn/99bxfAjSisF2DNnjv3r1J54vnoTJraDtO/PhgIQESIIHAElAaAxYSyDoBJRAYSnNjlCtXLp5+SP0jSXrcqFEjA2mb1DZe1ufi54DKkcNQwprjupRw5+etszq2yvtpW4MSdow333wzq/fJ9mB4X2bNmmUob1Hb/K3vm/qxYKi4htmeAscjARIggUIhQA1cYEXvYEwcdnHYmhs3bpx8+eWXnpOGMTmSiPfo0SMw21rYLkXYDauHI7YdlQCU90bxW7du1eFP/vjjj4Tng2eAWGn5WpDGDIF44TziVhCQGNvAXbp0cevGNhIgARIIFAEKcIF6XMGdLHJNzp07V8fh+uSTTzwXUqNGDR1AFdHvg2B/tWTJEmnfvr0tMCziwiE+HGzm8rWcf/758vjjjydM7+CDD9aOC1WrVk2oz4cTxKlDcN0XXnjBdTqHHXaYzpyB4NIHHniga182kgAJkEDQCNCJIWhPLKDzxRcohDF8+T722GNaY+W2FDhFwLgcgtw999zjatfkNk6u2mDzhlya1gIvVWSmyNeCAM1W4Q1zhV1fvglvu3bt0pk+oO10E94QfxCp3eCgAPYU3vL17eO8SIAEMiFADVwm9HhtgQkogwFZtGiRjnaPbUavcvjhh2sBaeDAgVK6dGmv7oXSji3Ipk2bOm7pIfhvvm3h4Rk0a9ZMBx82AytfvrwO2uvmUWzu7/cxtuEnTpwod955p8DZwq0gRh+2S8MUONptvWwjARKILgEKcNF99nmzctgyIW0R/noVZAQYMGCA1rDkY2BZaNwQHw5ekeZy6KGHyrp166RKlSrm6kI9njdvnlx44YW2OcBmsW/fvrb6XFdAwIS29uabbxbEG3QrJ510ktbUNm/e3K0b20iABEggNAQowIXmUQZ/IdDEIU0Xtsfw5e1WYKPVr18/GTx4cF4JRZgzsk8gF6y1tGzZUgupyLVZ2AVaLWTIsDqWIM/ru+++K4U9x1WrVmkHBeU16ooKoUCgmYMdX1BCtrguiI0kQAIkkCKBwv8mSXGi7BZ+Ath+XLhwod6C7N69u6sQgWTkiDWHrTIIcrB3ypeCNFtO26UrVqzQAmo+zBN2hVbhDfNCfWEKb3iO2AY95ZRTxE14gyYWiek//vhjHdeNwls+vFWcAwmQQC4JUAOXS9q8V1oEkDwdX9Jz5swReLG6lSJFikjXrl11UODjjjvOrWtO2hDEGMb227dvT7gf0jWtXLlS28olNOTw5Ntvv9VBe3/55ZeEu3bs2FEL0AmVOTpBAOiRI0fqrB5uzxr8oN28/fbbBWmwWEiABEggqgQowEX1yQdo3dAUIY7cjBkzBJo3twJNDASRIUOGCOyiCrNA49a6dWtRwX4TpgHPWsQugxapMAo0hLBzMxd4aiKNVu3atc3Vvh//+eefWmiD8OaVuePss8/W7wFCs7CQAAmQQNQJcAs16m9AANaPcBbYLkVoEcT/KlmyZNJZw3YO27BNmjSR008/XV577bWkff1ugM0bDPCtJRYixVqfi3M4UsycOdN2K4RsybXwBs/cY489VocscRPekNf0lVdekeeff14HHLZNnhUkQAIkEEEC1MBF8KEHfcn4sodAN2nSJMHWm1dBqAxo5Dp06ODVNevtf//9t7bncgqV8sgjj0jPnj2zfk+3ASHUQhgyF3jIwvYMgW9zUWDbhjhzXjlWK1WqJKNGjdKMCtMuLxdMeA8SIAESSJcABbh0ibF/3hCADde0adO04f0333zjOS94WN5yyy1y3nnn5dRQHxq3Bg0a2GKYQZP43nvv5SxmGTRY55xzjo2Tytmak2DDCAUCjSRCg7h5GSP+HDStSJOVzxksbCBZQQIkQAI5JEABLoeweSt/CCAkhkpoLmPHjhXk9fQqCJ+BfKuIgZarKP3JYq7BTg8hM/yeBzSB9erV016bZj61atWSDz/80Nf7I/guwsNAY4pnlaxAy3bJJZdoZwYEE2YhARIgARJIToA2cMnZsCUgBIoVK6bTbsFr9eGHH/a0k0LoCaT1gvBy//33uwoV2UKAxPAXX3yxbThsJw4bNsxWn+0KrBPrthY4h/glPEJoxH1jyeTdhLczzzxTO3Yghh6FN+tT4jkJkAAJ2AlQA2dnwpqAE4DX59NPP621Ptii9CoQGGCTBUN+P9NH7dmzR2CQ//nnnydMCZqnZcuWaY/VhIYsncBmEEKU1V6wVatWvjl5IBgzgiwjM4VbqVu3rowfP16QS5aFBEiABEggdQIU4FJnxZ4BJLB48WKdpgtJ270KjPivvvpq/R8M+/0o0Li1aNHCFtcOBvvwEEXO12wXJHZHLlFzgdD49ttva4HSXJ/pMQTmQYMGeaZFK1eunIwYMUKn7EIMPxYSIAESIIE0CShjYhYSCD0BFU7EOOOMM5Cfy/M/pYUzlPbIUI4RvnBRqZ8c59CpU6es3++TTz4x1Bap7X5qCzmr91IBi43evXsbSjC03cvMXKVAM5RHsKHs4rJ6fw5GAiRAAlEjQA1cmgIvuwebwNq1a7VGDrHi1D9218XAtg5J3bEViJyb2SrY4k0Wow42Y9jKzVb5z3/+I88++2zCcCVKlBDYC1aoUCGhviAne/fu1c4j2Ab99ddfkw6BAMsXXXSR3tauXLly0n5sIAESIAESSI0ABbjUOLFXyAgg6wCSoD/++OOyb98+19XByB8eqwiBka1gt1999ZUgrMn333+fcG+loRIImbANy7QsX75cTjvtNNsw2LocOnSorT6dCgih8Py99dZbBam53AoCGt99991ywgknuHVjGwmQAAmQQBoEKMClAYtdw0cAAWzvuusumT17tiCtk1uB3RhiyCGWHISvTMtzzz0nnTt3tg2DXK6wlYMGsKAFAtaJJ56o48yZx4D2S22rCgTFgpalS5dqO7cNGza4DgFhF6Fd1Nawaz82kgAJkAAJpE+AYUTSZ8YrQkQA3pkIXQHPUDgwuAk2EIoWLFigg/IiL+fq1aszIgHB5vLLL7eNAcEIgWwzKcjy4OSBC62j2xrd7ol4ce3atZO2bduKm/AGR4x7771Xx5ej8OZGlG0kQAIkUHAC1MAVnB2vDCGBXbt2CTITTJkyxZY5wWm52KJEmq42bdo4NXvW/fbbb1pTtnHjRltfhOIoSPov2KUhxp01O0Xjxo1lzZo1Anu0dMqOHTt0rLoZM2a4bjcXLVpUBg4cqHmULl06nVuwLwmQAAmQQJoEKMClCYzdo0Hgp59+kvvuu09nD/juu+88F92kSRO9tdqxY8e0BaT169cLMjL88ccfCfcpU6aMoC3dwLYIDDxy5MiEsXCycuVKHcLE1pCkAsLlPffco7eYEcPOrXTr1k3GjBkjNWrUcOvGNhIgARIggSwRoACXJZAcJpwEoM2aPn26Djb79ddfey4S9muwkYNAA5u5VMvkyZP1Fq61vwp9Ii+99FLKQqEK5yFIFWb1CIXt3hNPPGEd3vEc3rlz587VmrRt27Y59olVNmvWTAt5TZs2jVXxLwmQAAmQQA4IUIDLAWTeIvgEoB1Dmi44PCA5vVfBFibyrSJ9VqqpqmBXt2jRItvQSHeF4LiplJ49e8qcOXMSumJrc9OmTSlpx1asWKGTyL/zzjsJY1hPoGmDxg2CKgsJkAAJkEDuCVCAyz1z3jHABJDf87HHHtMhSJzs1qxLq1Klio4j169fP0/nAdjfHX/88bawHAcddJAgk4RXGA5kVsBWrDW+HeLYwRvUrcAzFY4T8Ix1K7BtQ+gQ2LphXiwkQAIkQAKFQ4ACXOFw510DTgBC0jPPPKMD03ppq7DUsmXL6nyrAwYMkJIlSyZdPXKiwsvTKoQhJMe7774rCMKbrJxyyimyatWqhGbY0SFob6lSpRLqYyeIQzd8+HCZNm2aLb1XrA/+QouIAMOwr/Mj3Zf5XjwmARIgARLwJpC6kY73WOxBApEhAE/OLl266HyiS5YsEQhPbmXnzp16S7Vq1apaCLIG8I1dC5u366+/PnYa/wsNGbReycqTTz5pE97QF8KZk/CGLWFkT0AYFdjf/fXXX8mG1rHqEPh40qRJFN6SUmIDCZAACeSWADVwueXNu4WYALw8R48erZ0OvJYJTRpiwEFYs6a0QkBhOAdA42YtyBxhtTtD/2OOOUY2b96c0P3YY4/VXqzWZPGIZQf7PC9bPmzZIoMCMimwkAAJkAAJ5BcBauDy63lwNgEmAC0ctHGwRYN2zi3eGrxbIRzBGQBbk1u3bo2vHLZl8+fPd9wuveyyy+SLL76I98UBNGNW4Q31GN8svCHwMATD7t27uwpvsNuDIwRSelF4A0kWEiABEsg/AtTA5d8z4YxCQgBODsh8AGHMK9/qAQccID169ND5VuvUqaMJzJw5U/r27Wuj0bx5c4G3KIQzxKjDNuju3bsT+sGODsIkCjRt0LhB8+ZWYJuHftddd52nw4XbOGwjARIgARLwn0CR21Xx/za8AwlEjwAcCKCJu+iii3SQXtiRJRPkkKZr3bp1MnXqVJ2mCmFI2rdvr8N/IIWVuSA2G2LMtWrVSocXsTouQLCDgwU0eUhaj1AmCAicrKD/pZdeqq8566yzUg57kmw81pMACZAACfhPgBo4/xnzDiSgCSAQMBwHEBgYW6heBXlHr7nmGnHaNoXQhXynvXr1sgmFEMbq1q0rI0aMkGTOErF74x6IM4f+LCRAAiRAAsEhQAEuOM+KMw0JAWx7Tpw4Uafqsm59Oi2xQYMGWoMGLZ25ICk90l2ZS/HixbVTxOeff26uth0j3hyESXi9spAACZAACQSPAAW44D0zzjgkBH7++WeZMmWKTJgwQRDENxcFHq/Ik9qnT5+0Un3lYm68BwmQAAmQQOoEKMClzoo9ScAXAshb+uCDD2qNGHKZ+lGgmUM6LmRbcAsG7Me9OSYJkAAJkED2CVCAyz5TjkgCBSKAeG6zZ8/WOUa9tkBTvQGcHZAfFfHpKlasmOpl7EcCJEACJJDnBCjA5fkD4vSiRwCeqgjYixAk8FwtaGndurWOBQcbOhYSIAESIIFwEaAAF67nydWEiADyoS5cuFBrzxBUN9WCrAxIXn/22Wenegn7kQAJkAAJBIwABbiAPTBON5oEkOQeMd3WrFmTFADiziH3af/+/QWBgVlIgARIgATCS4CptML7bLmyEBFAuA/kJnUqRYsW1RkUPvvsM52Wi8KbEyXWkQAJkEC4CFADF67nydWEmMDvv/+u48eNGjUqHggYAXgXLVok1apVC/HKuTQSIAESIAErAQpwViI8J4E8J7Bz507p1KmTTmAPb9VDDjkkz2fM6ZEACZAACWSbAAW4bBPleCSQIwJIx8WYbjmCzduQAAmQQJ4RoACXZw+E0yEBEiABEiABEiABLwJ0YvAixHYSIAESIAESIAESyDMCFODy7IFwOiRAAiRAAiRAAiTgRYACnBchtpMACZAACZAACZBAnhGgAJdnD4TTIQESIAESIAESIAEvAhTgvAixnQRIgARIgARIgATyjADz7eTZA+F0ckPgn3/+kR07duTmZryLJlCuXDnZf3/+ZuTrQAIkQALZIEABLhsUOUbgCEB4q1ixYuDmHeQJf/3111KhQoUgL4FzJwESIIG8IcCfw3nzKDgREiABEiABEiABEkiNADVwqXFirxATeO+99wTbeyzZJwBNZ8OGDbM/MEckARIggYgToAAX8ReAyxctvHFrj28CCZAACZBAkAhwCzVIT4tzJQESIAESIAESIAFFgAIcXwMSIAESIAESIAESCBgBCnABe2CcLgmQAAmQAAmQAAlQgOM7QAImAtu2bZN169aZahIPv/32W3njjTcSK7N8tnfvXnn99dflp59+yvLIHI4ESIAESCAsBCjAheVJch1ZITBr1iztNfnEE084jvfss89Ky5YtHduyVfnpp59KixYt5K233srWkByHBEiABEggZAQowIXsgXI5mRMwDEN69uzpu6Yt85l6j4CMEywkQAIkQALhI0ABLnzPlCvKkED58uXlxBNPlE6dOslnn32W4WiFd/mYMWNk06ZNhTcB3pkESIAESMA3AhTgfEPLgYNK4KCDDpLnnntODjvsMGnXrp189913SZeyZ88eueaaa/S2a9u2bWXlypW67yWXXCKTJk2KX7d+/XpB+y+//BKvGzBggIwdOzZ+nuxg4cKFcs455wi2b0899VQ9zpo1a+Ldp0+fruu7dOkic+bM0fWjRo2Sm2++Wfr27Stt2rSR1atXyzvvvCMdO3bUwumQIUME9nyxcv311wsEPtQ3atRIX4e1xcrGjRulW7duUq9ePT3ea6+9FmuSGTNm6Ps3adJExo8fL9BgspAACZAACfhLgAKcv3w5ekAJQHhbvHix7N69W2vifv/9d8eVQLB69dVX5fLLL5eaNWvKGWecIRB8jjnmGBkxYoT89ttv+rp7771Xli5dGhewPvjgA5k6daqcdNJJjuOaK7dv3y4vvPCCzJ07VwuLaINg+ffff8tHH32k792hQwc57rjj5NFHH5Vvvvkm7gBx8skna4GvbNmycsUVV+hcpF27dhXY+l111VXx20C4u/3226VEiRJy9dVX67led911un3r1q16nshYcf7552vBNubIcf/99+s5tW7dWrp37y7jxo2TefPmxcflAQmQAAmQgE8E1K9lFhKIHAGVWB1qIv0fjmNl+PDhRtWqVWOnhtJcGcWKFTOU0GMoezJDCV3GAQccoNvXrl2rr1++fLmhhDb9n9JQGfPnzzd+/PFHQwlDhtKOGT/88INx8MEHG0rAM+rWrauv7d+/v1G/fv34fcwHSlDS47700ku6esqUKcYhhxxi7Nu3T58r4Um3f/zxx8aWLVuM/fbbzxg8eLDx559/xodRGjPdRwmK8Tq1nRo/vummmwzMNVaUY4ah7P5ip4YS3uLzUxpGo2TJkobyjo23xw6qV69uoD22/qFDhxpKExhrNpJxjnfgAQmQAAmQQIEIMJWWT4Ixhw0HgaZNm2rNF7RWN954oxx55JHxhW3evFkf33bbbaKEKH18xBFHCEKRlC5dWnr37i3QvP3888+iBDB5+umnRQlt8swzz+gx77vvvvhYXgf777+/4D8UaPpQoGmDR+xDDz2ktWYvvviivketWrV0u/V/derU0dpCaM02bNgQn3Os34EHHhg7lKOOOkrPERVKUJTTTz9dihcvHm/HgRIo5csvv5Rly5YlhF6BFk99GtnGT7iYJyRAAiRAAhkRoACXET5eHAUC5557rt4aHDRokMDOK1awTYqitE7aLixWH/uLrUgIS0qrJ1deeaXe4oQdXK9evbRA16NHj1jXtP7GhEUISSiwt4OABc9Z2Lthy7NIkSK6zeyFeuedd2o7ugULFshjjz2mt1F1J4f/xe6BJjh1LFmyRG/ZKu1jvDfuAWERIU8eeOCBeD0PSIAESIAE/CdAGzj/GfMOISAAI38IYWbnAbUdqoWygQMHamHo7bfflsmTJ8cdE2rXri3t27cXBOaFjRwK+sJG7tJLLxW1NZsxGThbwFkC2jncCw4X0PgdfvjheuynnnpKOy8sWrRI28dBwKtSpYrWEkKD9uuvv3rO4eKLL9YODxA4YcentnYF46KgDo4TsJ/D+qFlvOCCCzzHZAcSIAESIIHMCFCAy4wfr44QAQhKZ599dnzFEJqwbVm5cmUt3DVu3FimTZumPVJjna699lp9TbVq1XQVnA+guYMHajZKmTJl9DYthLIJEyYItGzYvoUAB80gBEp4rkK4wj2xpQstILRnyj5N+vTp4zkNOCg8/PDDsmrVKn0tNJKxIMO4B4RTaOCw/n79+snRRx/tOSY7kAAJkAAJZEZgP1jOZTYEryaB4BGA/VjFihX1xCHIVKhQIaNFKAcCbROmnBVs40DjppwA4vXW83hDBgfKaUJKlSoVt5OLDfXXX3/pw5h92x9//CFFixbVddAMwl4tnZLsPhgDHruYg7lkm7N5bB6TAAmQQJQJ/GvQEmUKXDsJZEgAseOSFbPwhj7W82TXpVN/6KGHOnaPCW6xxpjwhvN0hTdck+w+aLMKb6hjIQESIAES8IcAt1D94cpRSYAESIAESIAESMA3AhTgfEPLgUmABEiABEiABEjAHwIU4PzhylFJgARIgARIgARIwDcCtIHzDS0HDgqBHTt2BGWqgZsn2QbukXHCJEACASFAAS4gD4rT9I9Aw4YN/RucI5MACZAACZCADwS4heoDVA5JAiRAAiRAAiRAAn4SYBw4P+ly7LwlgBRT3N7L7eMpV66cLU5dbmfAu5EACZBAeAhQgAvPs+RKSIAESIAESIAEIkKAW6gRedBcJgmQAAmQAAmQQHgIUIALz7PkSkiABEiABEiABCJCgAJcRB40l0kCJEACJEACJBAeAhTgwvMsuRISIAESIAESIIGIEKAAF5EHzWWSAAmQAAmQAAmEhwAFuPA8S66EBEiABEiABEggIgQowEXkQXOZJEACJEACJEAC4SFAAS48z5IrIQESIAESIAESiAgBCnARedBcJgmQAAmQAAmQQHgI/D9IlGxnf5Z+GwAAAABJRU5ErkJggg=="}}},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier\nfrom sklearn.neural_network import MLPClassifier","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train_val, X_test, y_train_val, y_test = train_test_split( mnist.data, mnist.target, test_size=10000, random_state=42)\nX_train, X_val, y_train, y_val = train_test_split( X_train_val, y_train_val, test_size=10000, random_state=42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"random_forest_clf = RandomForestClassifier(n_estimators=10, random_state=42)\nextra_trees_clf = ExtraTreesClassifier(n_estimators=10, random_state=42)\nmlp_clf = MLPClassifier(random_state=42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"estimators = [random_forest_clf, extra_trees_clf, mlp_clf]\nfor estimator in estimators:\n    print(\"Training the\", estimator)\n    estimator.fit(X_train, y_train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"[estimator.score(X_val, y_val) for estimator in estimators]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_val_predictions = np.empty((len(X_val), len(estimators)), dtype=np.float32)\n\nfor index, estimator in enumerate(estimators):\n    X_val_predictions[:, index] = estimator.predict(X_val)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_val_predictions","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"rnd_forest_blender = RandomForestClassifier(n_estimators=200, oob_score=True, random_state=42)\nrnd_forest_blender.fit(X_val_predictions, y_val)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"rnd_forest_blender.oob_score_","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"We can fine-tune this blender or try other types of blenders (e.g., an MLPClassifier), then select the best one using cross-validation, as always.\n\nWe have trained a blender, and together with the classifiers they form a stacking ensemble.\n\nLet's evaluate the ensemble on the test set. For each image in the test set, make predictions with all our classifiers, then feed the predictions to the blender to get the ensemble's predictions."},{"metadata":{"trusted":true},"cell_type":"code","source":"X_test_predictions = np.empty((len(X_test), len(estimators)), dtype=np.float32)\n\nfor index, estimator in enumerate(estimators):\n    X_test_predictions[:, index] = estimator.predict(X_test)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"y_pred = rnd_forest_blender.predict(X_test_predictions)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"accuracy_score(y_test, y_pred)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### End\nIf you reached this far please comment and upvote this kernel, feel free to make improvements on the kernel and please share if you found anything useful !"}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":1}