{
  "id": 10955,
  "title": "Teacher's Key",
  "url": "/competitions/seizure-prediction/discussion/10955",
  "author_name": "",
  "post_date": "2014-11-18T18:48:53.527Z",
  "votes": 6,
  "comment_count": 7,
  "views": 3804,
  "content": "<p>I am helping&nbsp;teach a course in Applied Topology, and would like to use this data set (or something&nbsp;similar) as an example to test ideas in Topological Data Analysis.&nbsp; Now that the competition is over, would it be possible to get the 'key' to score the results as was done for the competition?&nbsp; My hope is that we assign the exact competition, have the instructors determine the resulting score, and let students compare their results to the final leaderboard.</p>\n\n<p>What do you think?</p>",
  "messages": [
    {
      "id": "58333",
      "postDate": "11/18/2014 18:48:53",
      "content": "<p>I am helping&nbsp;teach a course in Applied Topology, and would like to use this data set (or something&nbsp;similar) as an example to test ideas in Topological Data Analysis.&nbsp; Now that the competition is over, would it be possible to get the 'key' to score the results as was done for the competition?&nbsp; My hope is that we assign the exact competition, have the instructors determine the resulting score, and let students compare their results to the final leaderboard.</p>\n\n<p>What do you think?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "58363",
      "postDate": "11/19/2014 03:36:21",
      "content": "<p>We do plan to post the answer key for this competition, and use of the data for education is certainly appropriate.&nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "58411",
      "postDate": "11/19/2014 14:08:08",
      "content": "<p>Hello,</p>\n<p>Would it be possible to get also the 'keys' for the test data set used for&nbsp;Seizure Detection challenge evaluation?</p>\n<p>I developed a method that generalizes an algorithm trained on pooled data, instead of training subject specific models, but the results I get on the iEEG records made available for that challenge are with 0,04 bigger than the ones I obtain on the private leaderboard (locally: 0.986644, LB: 0.94711). I'm working now on measuring the sensitivity, false alarms rate and median latency per subject and I'm really curious how does my method perform on the entire data.</p>\n<p>Thank you, <br>Narcisa PANTILIMON</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "59729",
      "postDate": "12/07/2014 10:19:25",
      "content": "<p>When will the Key be posted?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "63429",
      "postDate": "02/02/2015 10:20:29",
      "content": "<p>Hi</p>\n<p>Any updates on Teacher's Key? Can we use the data for publications and acknowledge the source?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "63489",
      "postDate": "02/03/2015 16:07:29",
      "content": "<p>Hi</p>\n\n<p>I also have the same question and concern. If it's decided to release answer key, when would it be exactly? And what's the policy to use this data for publications?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "82009",
      "postDate": "06/16/2015 15:56:04",
      "content": "<p>Dear Kagglers,</p>\n<p>Attached to this post is the answer key from the seizure prediction competition. I hope this is helpful as you continue to develop and refine your innovative approaches.</p>\n<p>When you use the data from the competition in publications or other ways we would be grateful if you would acknowledge the competition's sponsors and organizers, as well as the IEEG portal at ieeg.org.</p>\n<p>Thank you again for making this an excellent competition!</p>\n<p>Regards,</p>\n<p>Ben</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "136692",
      "postDate": "09/24/2016 00:44:41",
      "content": "<p>Hi Ben, </p>\n\n<p>I am reading the paper&quot;Crowdsourcing reproducible seizure forecasting in human and canine epilepsy&quot;\nand found that there are holdout data available for extra testing in <a href=\"http://msel.mayo.edu/data.html\">http://msel.mayo.edu/data.html</a>. </p>\n\n<p>Would it also be possible to get the 'key'  for the holdout data?</p>\n\n<p>Thanks,\nPenning</p>",
      "rawMarkdown": "Hi Ben, \r\n\r\nI am reading the paper\"Crowdsourcing reproducible seizure forecasting in human and canine epilepsy\"\r\nand found that there are holdout data available for extra testing in http://msel.mayo.edu/data.html. \r\n\r\nWould it also be possible to get the 'key'  for the holdout data?\r\n\r\nThanks,\r\nPenning",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 58363,
      "author_name": "bbrinkm",
      "author_url": "",
      "post_date": "11/19/2014 03:36:21",
      "content": "<p>We do plan to post the answer key for this competition, and use of the data for education is certainly appropriate.&nbsp;</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 58411,
      "author_name": "narcisapantilimon",
      "author_url": "",
      "post_date": "11/19/2014 14:08:08",
      "content": "<p>Hello,</p>\n<p>Would it be possible to get also the 'keys' for the test data set used for&nbsp;Seizure Detection challenge evaluation?</p>\n<p>I developed a method that generalizes an algorithm trained on pooled data, instead of training subject specific models, but the results I get on the iEEG records made available for that challenge are with 0,04 bigger than the ones I obtain on the private leaderboard (locally: 0.986644, LB: 0.94711). I'm working now on measuring the sensitivity, false alarms rate and median latency per subject and I'm really curious how does my method perform on the entire data.</p>\n<p>Thank you, <br>Narcisa PANTILIMON</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 59729,
      "author_name": "chuzzelwit",
      "author_url": "",
      "post_date": "12/07/2014 10:19:25",
      "content": "<p>When will the Key be posted?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 63429,
      "author_name": "geeklearner",
      "author_url": "",
      "post_date": "02/02/2015 10:20:29",
      "content": "<p>Hi</p>\n<p>Any updates on Teacher's Key? Can we use the data for publications and acknowledge the source?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 63489,
      "author_name": "farid4",
      "author_url": "",
      "post_date": "02/03/2015 16:07:29",
      "content": "<p>Hi</p>\n\n<p>I also have the same question and concern. If it's decided to release answer key, when would it be exactly? And what's the policy to use this data for publications?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 82009,
      "author_name": "bbrinkm",
      "author_url": "",
      "post_date": "06/16/2015 15:56:04",
      "content": "<p>Dear Kagglers,</p>\n<p>Attached to this post is the answer key from the seizure prediction competition. I hope this is helpful as you continue to develop and refine your innovative approaches.</p>\n<p>When you use the data from the competition in publications or other ways we would be grateful if you would acknowledge the competition's sponsors and organizers, as well as the IEEG portal at ieeg.org.</p>\n<p>Thank you again for making this an excellent competition!</p>\n<p>Regards,</p>\n<p>Ben</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 136692,
      "author_name": "penningyu",
      "author_url": "",
      "post_date": "09/24/2016 00:44:41",
      "content": "<p>Hi Ben, </p>\n\n<p>I am reading the paper&quot;Crowdsourcing reproducible seizure forecasting in human and canine epilepsy&quot;\nand found that there are holdout data available for extra testing in <a href=\"http://msel.mayo.edu/data.html\">http://msel.mayo.edu/data.html</a>. </p>\n\n<p>Would it also be possible to get the 'key'  for the holdout data?</p>\n\n<p>Thanks,\nPenning</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "58333": "",
    "58363": "",
    "58411": "",
    "59729": "",
    "63429": "",
    "63489": "",
    "82009": "",
    "136692": "Hi Ben, \r\n\r\nI am reading the paper\"Crowdsourcing reproducible seizure forecasting in human and canine epilepsy\"\r\nand found that there are holdout data available for extra testing in http://msel.mayo.edu/data.html. \r\n\r\nWould it also be possible to get the 'key'  for the holdout data?\r\n\r\nThanks,\r\nPenning"
  },
  "source": "meta"
}