{
  "id": 93049,
  "title": "Worrying Trend",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/93049",
  "author_name": "",
  "post_date": "2019-05-22T17:49:08.106936900Z",
  "votes": 23,
  "comment_count": 1,
  "views": 0,
  "content": "<p>This competition is the first one that we have had a significant trend in public kernels being released with <strong><em>private datasets</em></strong>.</p>\n\n<p>If we believe the goal of Kernels is to educate and inform competitors with useful information in terms of both the competition in particular and machine learning in general there is no way having private datasets aids in this goal.  We don't know what this private data contains or even how it was produced!   I never thought I would wish for blends to come back LOL</p>",
  "messages": [
    {
      "id": "535328",
      "postDate": "05/22/2019 17:49:08",
      "content": "<p>This competition is the first one that we have had a significant trend in public kernels being released with <strong><em>private datasets</em></strong>.</p>\n\n<p>If we believe the goal of Kernels is to educate and inform competitors with useful information in terms of both the competition in particular and machine learning in general there is no way having private datasets aids in this goal.  We don't know what this private data contains or even how it was produced!   I never thought I would wish for blends to come back LOL</p>",
      "rawMarkdown": "This competition is the first one that we have had a significant trend in public kernels being released with ***private datasets***.\n\nIf we believe the goal of Kernels is to educate and inform competitors with useful information in terms of both the competition in particular and machine learning in general there is no way having private datasets aids in this goal.  We don't know what this private data contains or even how it was produced!   I never thought I would wish for blends to come back LOL",
      "votes": null
    },
    {
      "id": "538587",
      "postDate": "05/28/2019 19:41:07",
      "content": "<p>Yes, and as you said:</p>\n\n<p>assuming that the kernel is public <strong>for the reason to educate</strong> than the input dataset or the way to that very specific dataset is super important.</p>",
      "rawMarkdown": "Yes, and as you said:\n\n assuming that the kernel is public **for the reason to educate** than the input dataset or the way to that very specific dataset is super important.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 538587,
      "author_name": "frtgnn",
      "author_url": "",
      "post_date": "05/28/2019 19:41:07",
      "content": "<p>Yes, and as you said:</p>\n\n<p>assuming that the kernel is public <strong>for the reason to educate</strong> than the input dataset or the way to that very specific dataset is super important.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "535328": "This competition is the first one that we have had a significant trend in public kernels being released with ***private datasets***.\n\nIf we believe the goal of Kernels is to educate and inform competitors with useful information in terms of both the competition in particular and machine learning in general there is no way having private datasets aids in this goal.  We don't know what this private data contains or even how it was produced!   I never thought I would wish for blends to come back LOL",
    "538587": "Yes, and as you said:\n\n assuming that the kernel is public **for the reason to educate** than the input dataset or the way to that very specific dataset is super important."
  },
  "source": "meta"
}