{
  "id": 11076,
  "title": "Loading and training data",
  "url": "/competitions/avazu-ctr-prediction/discussion/11076",
  "author_name": "Pavitrakumar P",
  "post_date": "2014-11-28T09:42:12.170000",
  "votes": 0,
  "comment_count": 0,
  "views": 658,
  "content": "<p>Hi,&nbsp;</p>\n<p>This is my first competition on Kaggle and I have some trouble loading and training the data in python.</p>\n<p>I have been trying the&nbsp;one-hot encoding method &nbsp;but the number&nbsp;of unique features on each categorical feature is too large to fit into my memory - even&nbsp;when using a sparse matrix.&nbsp;</p>\n<p>How can I efficiently load and train the data?</p>\n<p>Also, is it correct to fit each row of the training date to my classifier if I use one-hot encoding method?..</p>\n<p>For loading data I have been using DataFrame from the pandas library.</p>",
  "messages": [
    {
      "id": 59148,
      "postDate": "2014-11-28T09:42:12.170Z",
      "content": "<p>Hi,&nbsp;</p>\n<p>This is my first competition on Kaggle and I have some trouble loading and training the data in python.</p>\n<p>I have been trying the&nbsp;one-hot encoding method &nbsp;but the number&nbsp;of unique features on each categorical feature is too large to fit into my memory - even&nbsp;when using a sparse matrix.&nbsp;</p>\n<p>How can I efficiently load and train the data?</p>\n<p>Also, is it correct to fit each row of the training date to my classifier if I use one-hot encoding method?..</p>\n<p>For loading data I have been using DataFrame from the pandas library.</p>"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "59148": ""
  }
}