{
  "id": 413894,
  "title": "About future data of wavelength",
  "url": "/competitions/tlvmc-parkinsons-freezing-gait-prediction/discussion/413894",
  "author_name": "",
  "post_date": "2023-05-30T14:30:29.942445400Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I have a question.</p>\n<p>This Competition we are predicting somewhere \"'StartHesitation', 'Turn', 'Walking'\" in the whole wavelength, so it is ok to have future data in it, right?</p>\n<p>I would think that Window would involve the future when using tsflex and seglearn.</p>\n<p>How do you think ?</p>",
  "messages": [
    {
      "id": "2281033",
      "postDate": "05/30/2023 14:30:29",
      "content": "<p>I have a question.</p>\n<p>This Competition we are predicting somewhere \"'StartHesitation', 'Turn', 'Walking'\" in the whole wavelength, so it is ok to have future data in it, right?</p>\n<p>I would think that Window would involve the future when using tsflex and seglearn.</p>\n<p>How do you think ?</p>",
      "rawMarkdown": "I have a question.\n \nThis Competition we are predicting somewhere \"'StartHesitation', 'Turn', 'Walking'\" in the whole wavelength, so it is ok to have future data in it, right?\n\nI would think that Window would involve the future when using tsflex and seglearn.\n\nHow do you think ?",
      "votes": null
    },
    {
      "id": "2281568",
      "postDate": "05/31/2023 01:48:31",
      "content": "<p>Hey Chronobcelp,</p>\n<p>To my knowledge there is no rule against using future times to place a data point in context.  I think there is still value in being able to diagnose/monitor progression after the fact.</p>\n<p>That said, how amazing would it be to predict FoG events in real time -- before they happened?  Perhaps a device could administer a rhythmic vibration or tone to help prevent the FoG event via <a href=\"https://youtu.be/3-wrNhyVTNE?t=98\" target=\"_blank\">Cueing</a>.</p>",
      "rawMarkdown": "Hey Chronobcelp,\n\nTo my knowledge there is no rule against using future times to place a data point in context.  I think there is still value in being able to diagnose/monitor progression after the fact.\n\nThat said, how amazing would it be to predict FoG events in real time -- before they happened?  Perhaps a device could administer a rhythmic vibration or tone to help prevent the FoG event via [Cueing](https://youtu.be/3-wrNhyVTNE?t=98).",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2281568,
      "author_name": "austinhinkel",
      "author_url": "",
      "post_date": "05/31/2023 01:48:31",
      "content": "<p>Hey Chronobcelp,</p>\n<p>To my knowledge there is no rule against using future times to place a data point in context.  I think there is still value in being able to diagnose/monitor progression after the fact.</p>\n<p>That said, how amazing would it be to predict FoG events in real time -- before they happened?  Perhaps a device could administer a rhythmic vibration or tone to help prevent the FoG event via <a href=\"https://youtu.be/3-wrNhyVTNE?t=98\" target=\"_blank\">Cueing</a>.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2281033": "I have a question.\n \nThis Competition we are predicting somewhere \"'StartHesitation', 'Turn', 'Walking'\" in the whole wavelength, so it is ok to have future data in it, right?\n\nI would think that Window would involve the future when using tsflex and seglearn.\n\nHow do you think ?",
    "2281568": "Hey Chronobcelp,\n\nTo my knowledge there is no rule against using future times to place a data point in context.  I think there is still value in being able to diagnose/monitor progression after the fact.\n\nThat said, how amazing would it be to predict FoG events in real time -- before they happened?  Perhaps a device could administer a rhythmic vibration or tone to help prevent the FoG event via [Cueing](https://youtu.be/3-wrNhyVTNE?t=98)."
  },
  "source": "meta"
}