{
  "id": 171699,
  "title": "QUANTILE REGRESSION",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/171699",
  "author_name": "",
  "post_date": "2020-08-02T05:51:09.958561600Z",
  "votes": 2,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Cant actually understand why quantile regression is suitable for this problem.\nCould any one help.\nAlso some links to understnad the basics of implementation of quantile regression would help</p>",
  "messages": [
    {
      "id": "954814",
      "postDate": "08/02/2020 05:51:09",
      "content": "<p>Cant actually understand why quantile regression is suitable for this problem.\nCould any one help.\nAlso some links to understnad the basics of implementation of quantile regression would help</p>",
      "rawMarkdown": "Cant actually understand why quantile regression is suitable for this problem.\nCould any one help.\nAlso some links to understnad the basics of implementation of quantile regression would help",
      "votes": null
    },
    {
      "id": "955060",
      "postDate": "08/02/2020 09:37:52",
      "content": "<p>Quantile regression is useful to estimate the confidence (aka the distribution) of the target variabile. \nI've implement a simple notebook <a href=\"https://www.kaggle.com/abiolatti/deep-quantile-regression-in-keras\">here</a> some time ago in keras.\nKeep in mind then when you are estimating, for example, the 0.9-quantile you're searching for the value such that 90% of the observations are smaller.</p>",
      "rawMarkdown": "Quantile regression is useful to estimate the confidence (aka the distribution) of the target variabile. \nI've implement a simple notebook [here](https://www.kaggle.com/abiolatti/deep-quantile-regression-in-keras) some time ago in keras.\nKeep in mind then when you are estimating, for example, the 0.9-quantile you're searching for the value such that 90% of the observations are smaller.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 955060,
      "author_name": "abiolatti",
      "author_url": "",
      "post_date": "08/02/2020 09:37:52",
      "content": "<p>Quantile regression is useful to estimate the confidence (aka the distribution) of the target variabile. \nI've implement a simple notebook <a href=\"https://www.kaggle.com/abiolatti/deep-quantile-regression-in-keras\">here</a> some time ago in keras.\nKeep in mind then when you are estimating, for example, the 0.9-quantile you're searching for the value such that 90% of the observations are smaller.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "954814": "Cant actually understand why quantile regression is suitable for this problem.\nCould any one help.\nAlso some links to understnad the basics of implementation of quantile regression would help",
    "955060": "Quantile regression is useful to estimate the confidence (aka the distribution) of the target variabile. \nI've implement a simple notebook [here](https://www.kaggle.com/abiolatti/deep-quantile-regression-in-keras) some time ago in keras.\nKeep in mind then when you are estimating, for example, the 0.9-quantile you're searching for the value such that 90% of the observations are smaller."
  },
  "source": "meta"
}