{
  "id": 337460,
  "title": "AMEX Datasets as Pickles - 918 Features",
  "url": "/competitions/amex-default-prediction/discussion/337460",
  "author_name": "Jafar Bakhshaliyev",
  "post_date": "2022-07-16T06:31:46.815000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi Guys,</p>\n<p>I uploaded AMEX datasets as a pickle version here - <a href=\"https://www.kaggle.com/datasets/jafarbakhshaliyev/amex-pickle-datasets-918-features?utm_medium=social&amp;utm_campaign=kaggle-dataset-share&amp;utm_source=twitter\" target=\"_blank\">https://www.kaggle.com/datasets/jafarbakhshaliyev/amex-pickle-datasets-918-features?utm_medium=social&amp;utm_campaign=kaggle-dataset-share&amp;utm_source=twitter</a> . <br>\nThese datasets have 918 features and give better results. I am working on hyperparameter tuning to get better results with LGBM and will try stacking too. </p>\n<p>You can upload the pickles to your notebook as:</p>\n<p>with gzip.open(\"./train_extended.pkl\", 'rb') as tr:<br>\n    train_data = pickle.load(tr)</p>\n<p>I hope it will help you and new entrants. Good Luck!</p>",
  "messages": [
    {
      "id": 1857476,
      "postDate": "2022-07-16T06:31:46.817Z",
      "content": "<p>Hi Guys,</p>\n<p>I uploaded AMEX datasets as a pickle version here - <a href=\"https://www.kaggle.com/datasets/jafarbakhshaliyev/amex-pickle-datasets-918-features?utm_medium=social&amp;utm_campaign=kaggle-dataset-share&amp;utm_source=twitter\" target=\"_blank\">https://www.kaggle.com/datasets/jafarbakhshaliyev/amex-pickle-datasets-918-features?utm_medium=social&amp;utm_campaign=kaggle-dataset-share&amp;utm_source=twitter</a> . <br>\nThese datasets have 918 features and give better results. I am working on hyperparameter tuning to get better results with LGBM and will try stacking too. </p>\n<p>You can upload the pickles to your notebook as:</p>\n<p>with gzip.open(\"./train_extended.pkl\", 'rb') as tr:<br>\n    train_data = pickle.load(tr)</p>\n<p>I hope it will help you and new entrants. Good Luck!</p>",
      "rawMarkdown": "Hi Guys,\n\nI uploaded AMEX datasets as a pickle version here - https://www.kaggle.com/datasets/jafarbakhshaliyev/amex-pickle-datasets-918-features?utm_medium=social&utm_campaign=kaggle-dataset-share&utm_source=twitter . \nThese datasets have 918 features and give better results. I am working on hyperparameter tuning to get better results with LGBM and will try stacking too. \n\nYou can upload the pickles to your notebook as:\n\nwith gzip.open(\"./train_extended.pkl\", 'rb') as tr:\n    train_data = pickle.load(tr)\n\nI hope it will help you and new entrants. Good Luck!\n\n",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1857476": "Hi Guys,\n\nI uploaded AMEX datasets as a pickle version here - https://www.kaggle.com/datasets/jafarbakhshaliyev/amex-pickle-datasets-918-features?utm_medium=social&utm_campaign=kaggle-dataset-share&utm_source=twitter . \nThese datasets have 918 features and give better results. I am working on hyperparameter tuning to get better results with LGBM and will try stacking too. \n\nYou can upload the pickles to your notebook as:\n\nwith gzip.open(\"./train_extended.pkl\", 'rb') as tr:\n    train_data = pickle.load(tr)\n\nI hope it will help you and new entrants. Good Luck!\n\n"
  }
}