{
  "id": 552949,
  "title": "Question about the training R2-score",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/552949",
  "author_name": "",
  "post_date": "2024-12-22T16:57:42.749353500Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I received a high validation R2-score in training but when I submit it , I get a negative score, what is wrong? Overfitting?</p>",
  "messages": [
    {
      "id": "3078680",
      "postDate": "12/22/2024 16:57:42",
      "content": "<p>I received a high validation R2-score in training but when I submit it , I get a negative score, what is wrong? Overfitting?</p>",
      "rawMarkdown": "I received a high validation R2-score in training but when I submit it , I get a negative score, what is wrong? Overfitting?",
      "votes": null
    },
    {
      "id": "3078706",
      "postDate": "12/22/2024 17:47:12",
      "content": "<p>You can if your metric function is wrong and also try to use all data to train your model again, because these are time-series data.</p>",
      "rawMarkdown": "You can if your metric function is wrong and also try to use all data to train your model again, because these are time-series data.",
      "votes": null
    },
    {
      "id": "3080553",
      "postDate": "12/25/2024 11:55:46",
      "content": "<p>It happened to me too. In my case it was because I split the data into train and validation sets wrongly.</p>",
      "rawMarkdown": "It happened to me too. In my case it was because I split the data into train and validation sets wrongly.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3078706,
      "author_name": "i2nfinit3y",
      "author_url": "",
      "post_date": "12/22/2024 17:47:12",
      "content": "<p>You can if your metric function is wrong and also try to use all data to train your model again, because these are time-series data.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3080553,
      "author_name": "max6296",
      "author_url": "",
      "post_date": "12/25/2024 11:55:46",
      "content": "<p>It happened to me too. In my case it was because I split the data into train and validation sets wrongly.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3078680": "I received a high validation R2-score in training but when I submit it , I get a negative score, what is wrong? Overfitting?",
    "3078706": "You can if your metric function is wrong and also try to use all data to train your model again, because these are time-series data.",
    "3080553": "It happened to me too. In my case it was because I split the data into train and validation sets wrongly."
  },
  "source": "meta"
}