{
  "id": 507281,
  "title": "How to handle the outliers correctly?",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/507281",
  "author_name": "AustinCKR",
  "post_date": "2024-05-25T07:26:58.629000",
  "votes": 2,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello, I was trying to remove the outliers from the dataset. However, I found that once I removed the outliers, even with the very loose restriction, the remaining dataset was much smaller. Then, my model becomes very underfit due to the lack of the sample to train. Hence, how do I properly handle the outliers in the dataset?</p>",
  "messages": [
    {
      "id": 2835159,
      "postDate": "2024-05-25T07:26:58.630Z",
      "content": "<p>Hello, I was trying to remove the outliers from the dataset. However, I found that once I removed the outliers, even with the very loose restriction, the remaining dataset was much smaller. Then, my model becomes very underfit due to the lack of the sample to train. Hence, how do I properly handle the outliers in the dataset?</p>",
      "rawMarkdown": "Hello, I was trying to remove the outliers from the dataset. However, I found that once I removed the outliers, even with the very loose restriction, the remaining dataset was much smaller. Then, my model becomes very underfit due to the lack of the sample to train. Hence, how do I properly handle the outliers in the dataset?",
      "votes": 2
    },
    {
      "id": 2835167,
      "postDate": "2024-05-25T07:32:42.727Z",
      "content": "<p>For outliers, I suggest following the link <a href=\"https://www.analyticsvidhya.com/blog/2021/05/detecting-and-treating-outliers-treating-the-odd-one-out/\" target=\"_blank\">here</a>. I suggest replacing them with median or mean values as it is easier to understand. It really depends on how you filter your data and what your model is</p>",
      "rawMarkdown": "For outliers, I suggest following the link [here](https://www.analyticsvidhya.com/blog/2021/05/detecting-and-treating-outliers-treating-the-odd-one-out/). I suggest replacing them with median or mean values as it is easier to understand. It really depends on how you filter your data and what your model is"
    },
    {
      "id": 2835177,
      "postDate": "2024-05-25T07:38:20.687Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 2835167,
      "author_name": "Corguin",
      "author_url": "",
      "post_date": "2024-05-25T07:32:42.727000",
      "content": "<p>For outliers, I suggest following the link <a href=\"https://www.analyticsvidhya.com/blog/2021/05/detecting-and-treating-outliers-treating-the-odd-one-out/\" target=\"_blank\">here</a>. I suggest replacing them with median or mean values as it is easier to understand. It really depends on how you filter your data and what your model is</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2835177,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-05-25T07:38:20.687000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2835159": "Hello, I was trying to remove the outliers from the dataset. However, I found that once I removed the outliers, even with the very loose restriction, the remaining dataset was much smaller. Then, my model becomes very underfit due to the lack of the sample to train. Hence, how do I properly handle the outliers in the dataset?",
    "2835167": "For outliers, I suggest following the link [here](https://www.analyticsvidhya.com/blog/2021/05/detecting-and-treating-outliers-treating-the-odd-one-out/). I suggest replacing them with median or mean values as it is easier to understand. It really depends on how you filter your data and what your model is",
    "2835177": ""
  }
}