{
  "id": 11121,
  "title": "features",
  "url": "/competitions/inria-bci-challenge/discussion/11121",
  "author_name": "",
  "post_date": "2014-12-03T12:38:36.757Z",
  "votes": null,
  "comment_count": 1,
  "views": 1051,
  "content": "<p>what futures do you use? (i don't know if this question is legal)</p>",
  "messages": [
    {
      "id": "59456",
      "postDate": "12/03/2014 12:38:36",
      "content": "<p>what futures do you use? (i don't know if this question is legal)</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "59491",
      "postDate": "12/03/2014 19:36:18",
      "content": "<p>If you look at phalaris' GBM benchmark code he uses 260 samples after the feedback signal, which correspond to the time when the letter that was selected is displayed on the screen. You could also try applying different transformations to the EEG data.&nbsp;Check out his post&nbsp;<a href=\"http://www.kaggle.com/c/inria-bci-challenge/forums/t/11056/beating-the-benchmark-with-gbm/59097#post59097\">here</a>.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 59491,
      "author_name": "tdevries",
      "author_url": "",
      "post_date": "12/03/2014 19:36:18",
      "content": "<p>If you look at phalaris' GBM benchmark code he uses 260 samples after the feedback signal, which correspond to the time when the letter that was selected is displayed on the screen. You could also try applying different transformations to the EEG data.&nbsp;Check out his post&nbsp;<a href=\"http://www.kaggle.com/c/inria-bci-challenge/forums/t/11056/beating-the-benchmark-with-gbm/59097#post59097\">here</a>.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "59456": "",
    "59491": ""
  },
  "source": "meta"
}