{
  "id": 14385,
  "title": "Wrappers for Feature selection",
  "url": "/competitions/facebook-recruiting-iv-human-or-bot/discussion/14385",
  "author_name": "",
  "post_date": "2015-05-25T11:00:59.413Z",
  "votes": -2,
  "comment_count": 2,
  "views": 1169,
  "content": "<p>Hi,</p>\n<p>I am looking for a feature selection (wrapper) on R (or Python). Currently I use a very manual process to select the right set of variables. It will be great if someone can share me codes/links which will help me automate this process to some extent.</p>",
  "messages": [
    {
      "id": "79982",
      "postDate": "05/25/2015 11:00:59",
      "content": "<p>Hi,</p>\n<p>I am looking for a feature selection (wrapper) on R (or Python). Currently I use a very manual process to select the right set of variables. It will be great if someone can share me codes/links which will help me automate this process to some extent.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "79983",
      "postDate": "05/25/2015 11:14:39",
      "content": "<p>Hi @Tavish, <br>In R, I would recommend the Boruta package or random forest itself.<br>cran.r-project.org/package=Boruta</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "80062",
      "postDate": "05/26/2015 18:45:13",
      "content": "<p>A vif criteria could be a good place to start specifically to eliminate multicolliniarty issue.&nbsp; Vif stands for variance inflation factor</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 79983,
      "author_name": "",
      "author_url": "",
      "post_date": "05/25/2015 11:14:39",
      "content": "<p>Hi @Tavish, <br>In R, I would recommend the Boruta package or random forest itself.<br>cran.r-project.org/package=Boruta</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 80062,
      "author_name": "daretoshare",
      "author_url": "",
      "post_date": "05/26/2015 18:45:13",
      "content": "<p>A vif criteria could be a good place to start specifically to eliminate multicolliniarty issue.&nbsp; Vif stands for variance inflation factor</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "79982": "",
    "79983": "",
    "80062": ""
  },
  "source": "meta"
}