{
  "id": 547226,
  "title": "Any evidence for COVID-19 market impact in data?",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/547226",
  "author_name": "Michael Timbs",
  "post_date": "2024-11-20T11:39:50.577000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Saw some strange behavior in CV today and it occurred to me that it is possible that the market response to COVID-19 could very well be in the training data somewhere. Has anyone found any evidence of serious baseline deviations or regime changes that may be explained by this?</p>\n<p>I have not done any work investigating this yet (because my CPU bound machine is churning data at 100% 24/7 and still training) but I'm curious if anyone else has had this thought?</p>",
  "messages": [
    {
      "id": 3050622,
      "postDate": "2024-11-20T11:39:50.577Z",
      "content": "<p>Saw some strange behavior in CV today and it occurred to me that it is possible that the market response to COVID-19 could very well be in the training data somewhere. Has anyone found any evidence of serious baseline deviations or regime changes that may be explained by this?</p>\n<p>I have not done any work investigating this yet (because my CPU bound machine is churning data at 100% 24/7 and still training) but I'm curious if anyone else has had this thought?</p>",
      "rawMarkdown": "Saw some strange behavior in CV today and it occurred to me that it is possible that the market response to COVID-19 could very well be in the training data somewhere. Has anyone found any evidence of serious baseline deviations or regime changes that may be explained by this?\n\nI have not done any work investigating this yet (because my CPU bound machine is churning data at 100% 24/7 and still training) but I'm curious if anyone else has had this thought?",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3050622": "Saw some strange behavior in CV today and it occurred to me that it is possible that the market response to COVID-19 could very well be in the training data somewhere. Has anyone found any evidence of serious baseline deviations or regime changes that may be explained by this?\n\nI have not done any work investigating this yet (because my CPU bound machine is churning data at 100% 24/7 and still training) but I'm curious if anyone else has had this thought?"
  }
}