{
  "id": 21079,
  "title": "Anyone trying to use xgboost?",
  "url": "/competitions/facebook-v-predicting-check-ins/discussion/21079",
  "author_name": "",
  "post_date": "2016-05-20T00:28:50.490Z",
  "votes": null,
  "comment_count": 1,
  "views": 593,
  "content": "<p>I'm a newbie of this kind of competition. </p>\n\n<p>For this classification task, I tried to use xgboost for entire training set, but it could not run the training (it got down and not responding). It seems like, too many instances and classes made it terminated. However, my machine is i7 and 16gb memory, so it looks like enough to fit whole dataset into memory.</p>\n\n<p>Is there anyone who has similar issue with me? How can I make it possible to train with xgboost?</p>",
  "messages": [
    {
      "id": "120694",
      "postDate": "05/20/2016 00:28:50",
      "content": "<p>I'm a newbie of this kind of competition. </p>\n\n<p>For this classification task, I tried to use xgboost for entire training set, but it could not run the training (it got down and not responding). It seems like, too many instances and classes made it terminated. However, my machine is i7 and 16gb memory, so it looks like enough to fit whole dataset into memory.</p>\n\n<p>Is there anyone who has similar issue with me? How can I make it possible to train with xgboost?</p>",
      "rawMarkdown": "I'm a newbie of this kind of competition. \r\n\r\nFor this classification task, I tried to use xgboost for entire training set, but it could not run the training (it got down and not responding). It seems like, too many instances and classes made it terminated. However, my machine is i7 and 16gb memory, so it looks like enough to fit whole dataset into memory.\r\n\r\nIs there anyone who has similar issue with me? How can I make it possible to train with xgboost?",
      "votes": null
    },
    {
      "id": "120700",
      "postDate": "05/20/2016 00:57:00",
      "content": "<p>See <a href=\"https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/21003/aws-usage\">https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/21003/aws-usage</a></p>",
      "rawMarkdown": "See https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/21003/aws-usage",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 120700,
      "author_name": "breakfastpirate",
      "author_url": "",
      "post_date": "05/20/2016 00:57:00",
      "content": "<p>See <a href=\"https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/21003/aws-usage\">https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/21003/aws-usage</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "120694": "I'm a newbie of this kind of competition. \r\n\r\nFor this classification task, I tried to use xgboost for entire training set, but it could not run the training (it got down and not responding). It seems like, too many instances and classes made it terminated. However, my machine is i7 and 16gb memory, so it looks like enough to fit whole dataset into memory.\r\n\r\nIs there anyone who has similar issue with me? How can I make it possible to train with xgboost?",
    "120700": "See https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/21003/aws-usage"
  },
  "source": "meta"
}