{
  "id": 66064,
  "title": "Features and cv stratified fold assignment",
  "url": "/competitions/rsna-pneumonia-detection-challenge/discussion/66064",
  "author_name": "Virilo Tejedor Aguilera",
  "post_date": "2018-09-17T21:32:14.914000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Regarding cross validation, I did <a href=\"https://www.kaggle.com/virilo/features-and-cv-stratified-fold-assignment\">this kernel</a> trying to distribute different cases and features among folds.</p>\n\n<p>I think it could be useful and little bit better than random split.</p>\n\n<p>It can be used also for simple train/val split</p>\n\n<p>Cheers,</p>\n\n<p>Virilo</p>",
  "messages": [
    {
      "id": 388944,
      "postDate": "2018-09-17T21:32:14.913Z",
      "content": "<p>Regarding cross validation, I did <a href=\"https://www.kaggle.com/virilo/features-and-cv-stratified-fold-assignment\">this kernel</a> trying to distribute different cases and features among folds.</p>\n\n<p>I think it could be useful and little bit better than random split.</p>\n\n<p>It can be used also for simple train/val split</p>\n\n<p>Cheers,</p>\n\n<p>Virilo</p>",
      "rawMarkdown": "Regarding cross validation, I did [this kernel][1] trying to distribute different cases and features among folds.\n\nI think it could be useful and little bit better than random split.\n\nIt can be used also for simple train/val split\n\nCheers,\n\nVirilo\n\n\n  [1]: https://www.kaggle.com/virilo/features-and-cv-stratified-fold-assignment",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "388944": "Regarding cross validation, I did [this kernel][1] trying to distribute different cases and features among folds.\n\nI think it could be useful and little bit better than random split.\n\nIt can be used also for simple train/val split\n\nCheers,\n\nVirilo\n\n\n  [1]: https://www.kaggle.com/virilo/features-and-cv-stratified-fold-assignment"
  }
}