{
  "id": 702,
  "title": "Gini vs AUC",
  "url": "/competitions/ClaimPredictionChallenge/discussion/702",
  "author_name": "",
  "post_date": "2011-07-13T05:07:28.120Z",
  "votes": null,
  "comment_count": 3,
  "views": 9445,
  "content": "<p>I know I could probably look this up, but what is the practical difference (if any) between this and the AUC method used in the overfit competition.</p>",
  "messages": [
    {
      "id": "4563",
      "postDate": "07/13/2011 05:07:28",
      "content": "<p>I know I could probably look this up, but what is the practical difference (if any) between this and the AUC method used in the overfit competition.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "4564",
      "postDate": "07/13/2011 05:19:32",
      "content": "<p>It's a bit more intuitive to think of the predicted vs actuals in terms of Gini where 0 is roughly equivalent to a random baseline. You're correct that they're similar given the identity</p>\r\n<p>\\[ G &#43; 1 = 2 AUC \\]</p>\r\n<p>Keep in mind that Gini in this context also works for non-binary actual values.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "4565",
      "postDate": "07/13/2011 05:34:51",
      "content": "<p>Ah - ok - also - do you think we could get more data?</p>\r\n<p>cars &lt;- read.csv(&quot;kaggle/auto/train_set.csv&quot;)<br>\r\n<br>\r\n&gt; nrow(cars)<br>\r\n[1] 13184290<br>\r\n<br>\r\n&gt; object.size(cars)<br>\r\n2531630944 bytes</p>\r\n<p>I am not sure if I can draw any conclusions with only 13 million examples :)</p>\r\n<p>I just got my SSD and this is the first file I think I actually had to leave the room while it loaded....</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "4576",
      "postDate": "07/13/2011 20:00:10",
      "content": "<p>There is actually no practical difference between the two. Both rely on a model correctly ranking the cases, i.e., bad cases should be ranked lower than the good cases.</p>\r\n<p>On this note, AUC actually has an intuitive meaning, i.e, if you randomly pick a bad case and randomly pick a good case, the probability that the bad case is ranked lower than the good one is simply the AUC%.</p>\r\n<p>However, gini does not have such straight forward interpretation. (of course you can convert the GINI to AUC and get the probability I refered above). Gini, from the Italian, Gini Corrado was just an extension of the Pareto graph.</p>\r\n<p>In both the cases you are looking to achive highest possible AUC/Gini.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 4564,
      "author_name": "jeffmoser",
      "author_url": "",
      "post_date": "07/13/2011 05:19:32",
      "content": "<p>It's a bit more intuitive to think of the predicted vs actuals in terms of Gini where 0 is roughly equivalent to a random baseline. You're correct that they're similar given the identity</p>\r\n<p>\\[ G &#43; 1 = 2 AUC \\]</p>\r\n<p>Keep in mind that Gini in this context also works for non-binary actual values.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 4565,
      "author_name": "chrisraimondi",
      "author_url": "",
      "post_date": "07/13/2011 05:34:51",
      "content": "<p>Ah - ok - also - do you think we could get more data?</p>\r\n<p>cars &lt;- read.csv(&quot;kaggle/auto/train_set.csv&quot;)<br>\r\n<br>\r\n&gt; nrow(cars)<br>\r\n[1] 13184290<br>\r\n<br>\r\n&gt; object.size(cars)<br>\r\n2531630944 bytes</p>\r\n<p>I am not sure if I can draw any conclusions with only 13 million examples :)</p>\r\n<p>I just got my SSD and this is the first file I think I actually had to leave the room while it loaded....</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 4576,
      "author_name": "sashikanthdareddy",
      "author_url": "",
      "post_date": "07/13/2011 20:00:10",
      "content": "<p>There is actually no practical difference between the two. Both rely on a model correctly ranking the cases, i.e., bad cases should be ranked lower than the good cases.</p>\r\n<p>On this note, AUC actually has an intuitive meaning, i.e, if you randomly pick a bad case and randomly pick a good case, the probability that the bad case is ranked lower than the good one is simply the AUC%.</p>\r\n<p>However, gini does not have such straight forward interpretation. (of course you can convert the GINI to AUC and get the probability I refered above). Gini, from the Italian, Gini Corrado was just an extension of the Pareto graph.</p>\r\n<p>In both the cases you are looking to achive highest possible AUC/Gini.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "4563": "",
    "4564": "",
    "4565": "",
    "4576": ""
  },
  "source": "meta"
}