{
  "id": 507191,
  "title": "How to handle missing values?",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/507191",
  "author_name": "",
  "post_date": "2024-05-24T18:56:58.769207300Z",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I found that, there's a lot missing values in each dataset, so how can I handle them well? I saw some people is like find the mean for the null values in each column and after it meets certain threshold and remove whole column for that, is there any reason why doing that?</p>",
  "messages": [
    {
      "id": "2834493",
      "postDate": "05/24/2024 18:56:58",
      "content": "<p>I found that, there's a lot missing values in each dataset, so how can I handle them well? I saw some people is like find the mean for the null values in each column and after it meets certain threshold and remove whole column for that, is there any reason why doing that?</p>",
      "rawMarkdown": "I found that, there's a lot missing values in each dataset, so how can I handle them well? I saw some people is like find the mean for the null values in each column and after it meets certain threshold and remove whole column for that, is there any reason why doing that?",
      "votes": null
    },
    {
      "id": "2835158",
      "postDate": "05/25/2024 07:26:33",
      "content": "<p>Usually for missing values, it depends on your dataset. Go filter the columns first, see how many attributes you have remaining, then based on that, handle the missing values accordingly</p>",
      "rawMarkdown": "Usually for missing values, it depends on your dataset. Go filter the columns first, see how many attributes you have remaining, then based on that, handle the missing values accordingly",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2835158,
      "author_name": "corguin",
      "author_url": "",
      "post_date": "05/25/2024 07:26:33",
      "content": "<p>Usually for missing values, it depends on your dataset. Go filter the columns first, see how many attributes you have remaining, then based on that, handle the missing values accordingly</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2834493": "I found that, there's a lot missing values in each dataset, so how can I handle them well? I saw some people is like find the mean for the null values in each column and after it meets certain threshold and remove whole column for that, is there any reason why doing that?",
    "2835158": "Usually for missing values, it depends on your dataset. Go filter the columns first, see how many attributes you have remaining, then based on that, handle the missing values accordingly"
  },
  "source": "meta"
}