{
  "id": 9667,
  "title": "millions of training examples",
  "url": "/competitions/avito-prohibited-content/discussion/9667",
  "author_name": "",
  "post_date": "2014-07-07T20:02:48.700Z",
  "votes": null,
  "comment_count": 4,
  "views": 1510,
  "content": "<p>Training data sets have been getting larger and larger, haven't they? Is it really the case that with enormous amounts of training data, the resulting methodologies will be substantially different or better? Certainly, with more data you'll end up with more accurate models, but are the modelling techniques any better, and is this good for the competition and for the interests of the organizers?</p>",
  "messages": [
    {
      "id": "50108",
      "postDate": "07/07/2014 20:02:48",
      "content": "<p>Training data sets have been getting larger and larger, haven't they? Is it really the case that with enormous amounts of training data, the resulting methodologies will be substantially different or better? Certainly, with more data you'll end up with more accurate models, but are the modelling techniques any better, and is this good for the competition and for the interests of the organizers?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "50110",
      "postDate": "07/07/2014 21:31:00",
      "content": "<p>It is completely up to any participant to use all data or rather draw a sample. Also which data to use for training is an important question to answer. Sometimes removal of noise can lead to a better model accuracy.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "50111",
      "postDate": "07/07/2014 21:38:45",
      "content": "<p>[quote=Ivan Guz;50110]</p>\n<p>It is completely up to any participant to use all data or rather draw a sample.</p>\n<p>[/quote]</p>\n<p>Of course. But if you want to win, you should use all the data. Indeed, hardware resources are &nbsp;undoubtedly an edge in a competition with so much training data.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "50112",
      "postDate": "07/07/2014 21:41:02",
      "content": "",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "50147",
      "postDate": "07/08/2014 18:35:24",
      "content": "<p>@Jose I can see your point, but in this competition the data is mostly text, meaning it's high-dimensional, meaning you need many examples.</p>\n\n<p>Coming from the Criteo comp with 45m points, 4m here is a welcome breather :)</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 50110,
      "author_name": "ivanguz",
      "author_url": "",
      "post_date": "07/07/2014 21:31:00",
      "content": "<p>It is completely up to any participant to use all data or rather draw a sample. Also which data to use for training is an important question to answer. Sometimes removal of noise can lead to a better model accuracy.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 50111,
      "author_name": "solorzano",
      "author_url": "",
      "post_date": "07/07/2014 21:38:45",
      "content": "<p>[quote=Ivan Guz;50110]</p>\n<p>It is completely up to any participant to use all data or rather draw a sample.</p>\n<p>[/quote]</p>\n<p>Of course. But if you want to win, you should use all the data. Indeed, hardware resources are &nbsp;undoubtedly an edge in a competition with so much training data.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 50112,
      "author_name": "sirguessalot",
      "author_url": "",
      "post_date": "07/07/2014 21:41:02",
      "content": "",
      "votes": null,
      "replies": []
    },
    {
      "id": 50147,
      "author_name": "zygmunt",
      "author_url": "",
      "post_date": "07/08/2014 18:35:24",
      "content": "<p>@Jose I can see your point, but in this competition the data is mostly text, meaning it's high-dimensional, meaning you need many examples.</p>\n\n<p>Coming from the Criteo comp with 45m points, 4m here is a welcome breather :)</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "50108": "",
    "50110": "",
    "50111": "",
    "50112": "",
    "50147": ""
  },
  "source": "meta"
}