{
  "id": 61186,
  "title": "Bronze solution with github link",
  "url": "/competitions/avito-demand-prediction/writeups/new-ods-ai-group-in-kaggle-please-be-gentle-we-tri",
  "author_name": "",
  "post_date": "2018-08-28T16:41:49.477Z",
  "votes": 3,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I with my team ranked 131st (TOP 7%) in this challenge. I know we had much to do, anyway I would like to share our solutions: <a href=\"https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)\">https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)</a></p>\n\n<p>Best single model achieved in the Public LB around 2210:\nSVD text features\nNima image quality features\nTfidf over desc and title\nCounts on title\nAgregated features\nRidge features</p>\n\n<p>One of the important lesson learned from the competition: \nJoin earlier and spent more times\nDon't use single ipynb file for competition\nLog each submission code and output\nTeammates are really help much to move further.\nStacking and assembling rules kaggle for sure </p>",
  "messages": [
    {
      "id": "357157",
      "postDate": "07/15/2018 12:03:56",
      "content": "<p>I with my team ranked 131st (TOP 7%) in this challenge. I know we had much to do, anyway I would like to share our solutions: <a href=\"https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)\">https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)</a></p>\n\n<p>Best single model achieved in the Public LB around 2210:\nSVD text features\nNima image quality features\nTfidf over desc and title\nCounts on title\nAgregated features\nRidge features</p>\n\n<p>One of the important lesson learned from the competition: \nJoin earlier and spent more times\nDon't use single ipynb file for competition\nLog each submission code and output\nTeammates are really help much to move further.\nStacking and assembling rules kaggle for sure </p>",
      "rawMarkdown": "I with my team ranked 131st (TOP 7%) in this challenge. I know we had much to do, anyway I would like to share our solutions: https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)\n\nBest single model achieved in the Public LB around 2210:\nSVD text features\nNima image quality features\nTfidf over desc and title\nCounts on title\nAgregated features\nRidge features\n\n\nOne of the important lesson learned from the competition: \nJoin earlier and spent more times\nDon't use single ipynb file for competition\nLog each submission code and output\nTeammates are really help much to move further.\nStacking and assembling rules kaggle for sure",
      "votes": null
    },
    {
      "id": "364142",
      "postDate": "07/30/2018 20:58:40",
      "content": "<p>Hi, thanks for sharing your solution.\nI was trying to repro it and seeing: </p>\n\n<pre><code>IOError: File train_img_features_v1.csv does not exist\n</code></pre>\n\n<p>Do I need to run <a href=\"https://github.com/Diyago/Machine-Learning-scripts/blob/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/nn_image_features.py\">https://github.com/Diyago/Machine-Learning-scripts/blob/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/nn_image_features.py</a> to produce the file?</p>\n\n<p>Thanks!</p>",
      "rawMarkdown": "Hi, thanks for sharing your solution.\nI was trying to repro it and seeing: \n\n    IOError: File train_img_features_v1.csv does not exist\n\nDo I need to run https://github.com/Diyago/Machine-Learning-scripts/blob/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/nn_image_features.py to produce the file?\n\nThanks!",
      "votes": null
    },
    {
      "id": "366678",
      "postDate": "08/06/2018 07:42:06",
      "content": "<p>Sorry for the late answer. You have to extract it this way: <a href=\"https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/neural-image-assessment\">https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/neural-image-assessment</a> - NIMA features</p>\n\n<p>Unfortunately I didn't use classical features. </p>",
      "rawMarkdown": "Sorry for the late answer. You have to extract it this way: https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/neural-image-assessment - NIMA features\n\nUnfortunately I didn't use classical features.",
      "votes": null
    },
    {
      "id": "366741",
      "postDate": "08/06/2018 11:34:34",
      "content": "<p>Thank you and no problem at all!</p>",
      "rawMarkdown": "Thank you and no problem at all!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 364142,
      "author_name": "dmitrykan",
      "author_url": "",
      "post_date": "07/30/2018 20:58:40",
      "content": "<p>Hi, thanks for sharing your solution.\nI was trying to repro it and seeing: </p>\n\n<pre><code>IOError: File train_img_features_v1.csv does not exist\n</code></pre>\n\n<p>Do I need to run <a href=\"https://github.com/Diyago/Machine-Learning-scripts/blob/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/nn_image_features.py\">https://github.com/Diyago/Machine-Learning-scripts/blob/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/nn_image_features.py</a> to produce the file?</p>\n\n<p>Thanks!</p>",
      "votes": null,
      "replies": [
        {
          "id": 366678,
          "author_name": "insaff",
          "author_url": "",
          "post_date": "08/06/2018 07:42:06",
          "content": "<p>Sorry for the late answer. You have to extract it this way: <a href=\"https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/neural-image-assessment\">https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/neural-image-assessment</a> - NIMA features</p>\n\n<p>Unfortunately I didn't use classical features. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 366741,
          "author_name": "dmitrykan",
          "author_url": "",
          "post_date": "08/06/2018 11:34:34",
          "content": "<p>Thank you and no problem at all!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "357157": "I with my team ranked 131st (TOP 7%) in this challenge. I know we had much to do, anyway I would like to share our solutions: https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)\n\nBest single model achieved in the Public LB around 2210:\nSVD text features\nNima image quality features\nTfidf over desc and title\nCounts on title\nAgregated features\nRidge features\n\n\nOne of the important lesson learned from the competition: \nJoin earlier and spent more times\nDon't use single ipynb file for competition\nLog each submission code and output\nTeammates are really help much to move further.\nStacking and assembling rules kaggle for sure",
    "364142": "Hi, thanks for sharing your solution.\nI was trying to repro it and seeing: \n\n    IOError: File train_img_features_v1.csv does not exist\n\nDo I need to run https://github.com/Diyago/Machine-Learning-scripts/blob/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/nn_image_features.py to produce the file?\n\nThanks!",
    "366678": "Sorry for the late answer. You have to extract it this way: https://github.com/Diyago/Machine-Learning-scripts/tree/master/DEEP%20LEARNING/Kaggle:%20Avito%20Demand%20Prediction%20Challenge%20(bronze%20solution)/image%20feat.%20extraction/neural-image-assessment - NIMA features\n\nUnfortunately I didn't use classical features.",
    "366741": "Thank you and no problem at all!"
  },
  "source": "meta"
}