{
  "id": 262686,
  "title": "The gap between CV and LB is too large",
  "url": "/competitions/seti-breakthrough-listen/discussion/262686",
  "author_name": "Aristarkh_BFG",
  "post_date": "2021-08-07T09:13:36.375000",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>My model reaches CV = 0.877, but on LB its score is 0.699.<br>\nHowever, another model with the exact same method of evaluation has CV = 0.82 and LB = 0.725. What could be the reason?</p>",
  "messages": [
    {
      "id": 1457687,
      "postDate": "2021-08-07T14:10:43.983Z",
      "content": "<p>I have experienced this too,  it could be due to differences between train and test datasets I think finding a way to overcome the cv/lb gap is the key to the competition. </p>",
      "rawMarkdown": "I have experienced this too,  it could be due to differences between train and test datasets I think finding a way to overcome the cv/lb gap is the key to the competition. ",
      "votes": 1
    },
    {
      "id": 1457157,
      "postDate": "2021-08-07T09:13:36.377Z",
      "content": "<p>My model reaches CV = 0.877, but on LB its score is 0.699.<br>\nHowever, another model with the exact same method of evaluation has CV = 0.82 and LB = 0.725. What could be the reason?</p>",
      "rawMarkdown": "My model reaches CV = 0.877, but on LB its score is 0.699.\nHowever, another model with the exact same method of evaluation has CV = 0.82 and LB = 0.725. What could be the reason?",
      "votes": 1
    },
    {
      "id": 1458246,
      "postDate": "2021-08-07T19:16:50.487Z",
      "content": "<p><a href=\"https://www.kaggle.com/c/seti-breakthrough-listen/discussion/256043\" target=\"_blank\">https://www.kaggle.com/c/seti-breakthrough-listen/discussion/256043</a></p>",
      "rawMarkdown": "https://www.kaggle.com/c/seti-breakthrough-listen/discussion/256043",
      "votes": 2
    }
  ],
  "comments": [
    {
      "id": 1457687,
      "author_name": "shiroe",
      "author_url": "",
      "post_date": "2021-08-07T14:10:43.983000",
      "content": "<p>I have experienced this too,  it could be due to differences between train and test datasets I think finding a way to overcome the cv/lb gap is the key to the competition. </p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1458246,
      "author_name": "Chris Deotte",
      "author_url": "",
      "post_date": "2021-08-07T19:16:50.487000",
      "content": "<p><a href=\"https://www.kaggle.com/c/seti-breakthrough-listen/discussion/256043\" target=\"_blank\">https://www.kaggle.com/c/seti-breakthrough-listen/discussion/256043</a></p>",
      "votes": 2,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1457687": "I have experienced this too,  it could be due to differences between train and test datasets I think finding a way to overcome the cv/lb gap is the key to the competition. ",
    "1457157": "My model reaches CV = 0.877, but on LB its score is 0.699.\nHowever, another model with the exact same method of evaluation has CV = 0.82 and LB = 0.725. What could be the reason?",
    "1458246": "https://www.kaggle.com/c/seti-breakthrough-listen/discussion/256043"
  }
}