{
  "id": 76403,
  "title": "Left skewness in target ",
  "url": "/competitions/elo-merchant-category-recommendation/discussion/76403",
  "author_name": "",
  "post_date": "2019-01-02T13:29:47.022907500Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>The target variable is LEFT skewed with skewness value of -6.72. And I studied somewhere that (linear) models love normally distributed data ,and we need to transform this variable and make it more normally distributed. I tried cube root but it didn't help. Any suggestion is helpful and please correct me if I'm wrong.</p>",
  "messages": [
    {
      "id": "448981",
      "postDate": "01/02/2019 13:29:47",
      "content": "<p>The target variable is LEFT skewed with skewness value of -6.72. And I studied somewhere that (linear) models love normally distributed data ,and we need to transform this variable and make it more normally distributed. I tried cube root but it didn't help. Any suggestion is helpful and please correct me if I'm wrong.</p>",
      "rawMarkdown": "The target variable is LEFT skewed with skewness value of -6.72. And I studied somewhere that (linear) models love normally distributed data ,and we need to transform this variable and make it more normally distributed. I tried cube root but it didn't help. Any suggestion is helpful and please correct me if I'm wrong.",
      "votes": null
    },
    {
      "id": "449061",
      "postDate": "01/02/2019 15:33:40",
      "content": "<p>Thanks for the insight, take a look at this Blog explanation:\n<a href=\"https://machinelearningmastery.com/how-to-transform-data-to-fit-the-normal-distribution/\">https://machinelearningmastery.com/how-to-transform-data-to-fit-the-normal-distribution/</a></p>",
      "rawMarkdown": "Thanks for the insight, take a look at this Blog explanation:\nhttps://machinelearningmastery.com/how-to-transform-data-to-fit-the-normal-distribution/",
      "votes": null
    },
    {
      "id": "453136",
      "postDate": "01/09/2019 18:17:33",
      "content": "<p>log transformation can help you in decreasing the skewness in the data</p>",
      "rawMarkdown": "log transformation can help you in decreasing the skewness in the data",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 449061,
      "author_name": "cv13j0",
      "author_url": "",
      "post_date": "01/02/2019 15:33:40",
      "content": "<p>Thanks for the insight, take a look at this Blog explanation:\n<a href=\"https://machinelearningmastery.com/how-to-transform-data-to-fit-the-normal-distribution/\">https://machinelearningmastery.com/how-to-transform-data-to-fit-the-normal-distribution/</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 453136,
      "author_name": "kalyankkr",
      "author_url": "",
      "post_date": "01/09/2019 18:17:33",
      "content": "<p>log transformation can help you in decreasing the skewness in the data</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "448981": "The target variable is LEFT skewed with skewness value of -6.72. And I studied somewhere that (linear) models love normally distributed data ,and we need to transform this variable and make it more normally distributed. I tried cube root but it didn't help. Any suggestion is helpful and please correct me if I'm wrong.",
    "449061": "Thanks for the insight, take a look at this Blog explanation:\nhttps://machinelearningmastery.com/how-to-transform-data-to-fit-the-normal-distribution/",
    "453136": "log transformation can help you in decreasing the skewness in the data"
  },
  "source": "meta"
}