{
  "id": 244701,
  "title": "Exploitation vs. Exploration",
  "url": "/competitions/coleridgeinitiative-show-us-the-data/discussion/244701",
  "author_name": "Ken Miller",
  "post_date": "2021-06-07T21:36:38.665000",
  "votes": 6,
  "comment_count": 0,
  "views": 0,
  "content": "<p>This kind of reminds me of a reinforcement learning problem in the sense that if you get too loose with your model, you are heavily penalized for false positives, but if you are too strict with your model, you miss a lot of datasets.</p>\n<p>Really hard to generalize on things that are not seen without several false positives creeping in.</p>",
  "messages": [
    {
      "id": 1340399,
      "postDate": "2021-06-07T21:36:38.667Z",
      "content": "<p>This kind of reminds me of a reinforcement learning problem in the sense that if you get too loose with your model, you are heavily penalized for false positives, but if you are too strict with your model, you miss a lot of datasets.</p>\n<p>Really hard to generalize on things that are not seen without several false positives creeping in.</p>",
      "rawMarkdown": "This kind of reminds me of a reinforcement learning problem in the sense that if you get too loose with your model, you are heavily penalized for false positives, but if you are too strict with your model, you miss a lot of datasets.\n\nReally hard to generalize on things that are not seen without several false positives creeping in.",
      "votes": 6
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1340399": "This kind of reminds me of a reinforcement learning problem in the sense that if you get too loose with your model, you are heavily penalized for false positives, but if you are too strict with your model, you miss a lot of datasets.\n\nReally hard to generalize on things that are not seen without several false positives creeping in."
  }
}