{
  "id": 552035,
  "title": "Is the weight in the dataset equal to \"sample_weight\"?",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/552035",
  "author_name": "",
  "post_date": "2024-12-17T09:29:09.247845900Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Some public notebook set the parameter \"sample_weight=w_train\" when train lgb or xgb model, but do the \"weight\" in dataset and sample_weight in the model mean the same thing?</p>",
  "messages": [
    {
      "id": "3074065",
      "postDate": "12/17/2024 09:29:09",
      "content": "<p>Some public notebook set the parameter \"sample_weight=w_train\" when train lgb or xgb model, but do the \"weight\" in dataset and sample_weight in the model mean the same thing?</p>",
      "rawMarkdown": "Some public notebook set the parameter \"sample_weight=w_train\" when train lgb or xgb model, but do the \"weight\" in dataset and sample_weight in the model mean the same thing?",
      "votes": null
    },
    {
      "id": "3074102",
      "postDate": "12/17/2024 10:22:29",
      "content": "<p>Yes, this is correct <a href=\"https://www.kaggle.com/i2nfinit3y\" target=\"_blank\">@i2nfinit3y</a> </p>",
      "rawMarkdown": "Yes, this is correct @i2nfinit3y",
      "votes": null
    },
    {
      "id": "3074161",
      "postDate": "12/17/2024 11:41:16",
      "content": "<p>But do the sample_weight works for imbalanced label distribution? And the weight data competition provide are just for metric</p>",
      "rawMarkdown": "But do the sample_weight works for imbalanced label distribution? And the weight data competition provide are just for metric",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3074102,
      "author_name": "ravi20076",
      "author_url": "",
      "post_date": "12/17/2024 10:22:29",
      "content": "<p>Yes, this is correct <a href=\"https://www.kaggle.com/i2nfinit3y\" target=\"_blank\">@i2nfinit3y</a> </p>",
      "votes": null,
      "replies": [
        {
          "id": 3074161,
          "author_name": "i2nfinit3y",
          "author_url": "",
          "post_date": "12/17/2024 11:41:16",
          "content": "<p>But do the sample_weight works for imbalanced label distribution? And the weight data competition provide are just for metric</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3074065": "Some public notebook set the parameter \"sample_weight=w_train\" when train lgb or xgb model, but do the \"weight\" in dataset and sample_weight in the model mean the same thing?",
    "3074102": "Yes, this is correct @i2nfinit3y",
    "3074161": "But do the sample_weight works for imbalanced label distribution? And the weight data competition provide are just for metric"
  },
  "source": "meta"
}