{
  "id": 201438,
  "title": "Tabnet in the ensemble",
  "url": "/competitions/riiid-test-answer-prediction/discussion/201438",
  "author_name": "Sahil Maheshwari",
  "post_date": "2020-12-05T03:11:57.228000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi , I had recently created ensemble of Catboost and LIghtGBM and couldn't see a huge improvement in the results , is the diversity in ensemble the trick for getting better lb , I am assuming tabnet must be more diverse ? What models are people using for ensemble ?</p>",
  "messages": [
    {
      "id": 1102531,
      "postDate": "2020-12-05T03:11:57.230Z",
      "content": "<p>Hi , I had recently created ensemble of Catboost and LIghtGBM and couldn't see a huge improvement in the results , is the diversity in ensemble the trick for getting better lb , I am assuming tabnet must be more diverse ? What models are people using for ensemble ?</p>",
      "rawMarkdown": "Hi , I had recently created ensemble of Catboost and LIghtGBM and couldn't see a huge improvement in the results , is the diversity in ensemble the trick for getting better lb , I am assuming tabnet must be more diverse ? What models are people using for ensemble ?"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1102531": "Hi , I had recently created ensemble of Catboost and LIghtGBM and couldn't see a huge improvement in the results , is the diversity in ensemble the trick for getting better lb , I am assuming tabnet must be more diverse ? What models are people using for ensemble ?"
  }
}