{
  "id": 16597,
  "title": "Top of the LB. Beat the leak?",
  "url": "/competitions/dato-native/discussion/16597",
  "author_name": "",
  "post_date": "2015-09-21T04:03:15.133Z",
  "votes": null,
  "comment_count": 2,
  "views": 612,
  "content": "<p>Hey y'all up at the top of the LB. Do any of you think that by the end of this you'll manage to beat the leak? I'm just curious...</p>",
  "messages": [
    {
      "id": "93064",
      "postDate": "09/21/2015 04:03:15",
      "content": "<p>Hey y'all up at the top of the LB. Do any of you think that by the end of this you'll manage to beat the leak? I'm just curious...</p>",
      "rawMarkdown": "Hey y'all up at the top of the LB. Do any of you think that by the end of this you'll manage to beat the leak? I'm just curious...",
      "votes": null
    },
    {
      "id": "93095",
      "postDate": "09/21/2015 13:57:07",
      "content": "<p>I'm fairly certain I've seen misclassifications in the training data set, but the definition of native advertising isn't precise enough to be sure.  I suspect errors in the data set will limit us to below 0.999 AUROC.</p>\n\n<p><strong>eta:</strong> Correction: By definition, there can't be misclassifications in the data, due to the definition of &quot;native ads&quot; used in this competition.</p>",
      "rawMarkdown": "I'm fairly certain I've seen misclassifications in the training data set, but the definition of native advertising isn't precise enough to be sure.  I suspect errors in the data set will limit us to below 0.999 AUROC.\r\n\r\n**eta:** Correction: By definition, there can't be misclassifications in the data, due to the definition of \"native ads\" used in this competition.",
      "votes": null
    },
    {
      "id": "93097",
      "postDate": "09/21/2015 14:23:59",
      "content": "<p>Well, since the leak gave .99805, I'd say that would do it! I just think that's a cute idea. </p>\n\n<p>Anyway, I'm just hoping to make it to .98 by the time this is up.</p>",
      "rawMarkdown": "Well, since the leak gave .99805, I'd say that would do it! I just think that's a cute idea. \r\n\r\nAnyway, I'm just hoping to make it to .98 by the time this is up.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 93095,
      "author_name": "mortehu",
      "author_url": "",
      "post_date": "09/21/2015 13:57:07",
      "content": "<p>I'm fairly certain I've seen misclassifications in the training data set, but the definition of native advertising isn't precise enough to be sure.  I suspect errors in the data set will limit us to below 0.999 AUROC.</p>\n\n<p><strong>eta:</strong> Correction: By definition, there can't be misclassifications in the data, due to the definition of &quot;native ads&quot; used in this competition.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93097,
      "author_name": "gbrello",
      "author_url": "",
      "post_date": "09/21/2015 14:23:59",
      "content": "<p>Well, since the leak gave .99805, I'd say that would do it! I just think that's a cute idea. </p>\n\n<p>Anyway, I'm just hoping to make it to .98 by the time this is up.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "93064": "Hey y'all up at the top of the LB. Do any of you think that by the end of this you'll manage to beat the leak? I'm just curious...",
    "93095": "I'm fairly certain I've seen misclassifications in the training data set, but the definition of native advertising isn't precise enough to be sure.  I suspect errors in the data set will limit us to below 0.999 AUROC.\r\n\r\n**eta:** Correction: By definition, there can't be misclassifications in the data, due to the definition of \"native ads\" used in this competition.",
    "93097": "Well, since the leak gave .99805, I'd say that would do it! I just think that's a cute idea. \r\n\r\nAnyway, I'm just hoping to make it to .98 by the time this is up."
  },
  "source": "meta"
}