{
  "id": 191003,
  "title": "evaluation metric",
  "url": "/competitions/lyft-motion-prediction-autonomous-vehicles/discussion/191003",
  "author_name": "",
  "post_date": "2020-10-14T06:08:09.221932200Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Does \\(c^k\\) in the evaluation metric equation mean \\(c_k\\) instead?</p>",
  "messages": [
    {
      "id": "1049111",
      "postDate": "10/14/2020 06:08:09",
      "content": "<p>Does \\(c^k\\) in the evaluation metric equation mean \\(c_k\\) instead?</p>",
      "rawMarkdown": "Does \\\\(c^k\\\\) in the evaluation metric equation mean \\\\(c_k\\\\) instead?",
      "votes": null
    },
    {
      "id": "1057632",
      "postDate": "10/22/2020 20:05:42",
      "content": "<p>I believe in the evaluation metric, the superscript refers to which hypothesis, and the subscript refers to which time step.  The number \\(c\\) refers to the confidence.  You can only have up to three path hypotheses, so \\(k\\) can only be 0, 1, or 2.  You assign the confidence to the entire path, not to individual time steps.  This means there are a total of three confidences \\(c^k\\), and they are \\(c^0\\),\\( c^1\\), and \\(c^2\\).  Hope this helps.</p>",
      "rawMarkdown": "I believe in the evaluation metric, the superscript refers to which hypothesis, and the subscript refers to which time step.  The number \\\\(c\\\\) refers to the confidence.  You can only have up to three path hypotheses, so \\\\(k\\\\) can only be 0, 1, or 2.  You assign the confidence to the entire path, not to individual time steps.  This means there are a total of three confidences \\\\(c^k\\\\), and they are \\\\(c^0\\\\),\\\\( c^1\\\\), and \\\\(c^2\\\\).  Hope this helps.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1057632,
      "author_name": "socated",
      "author_url": "",
      "post_date": "10/22/2020 20:05:42",
      "content": "<p>I believe in the evaluation metric, the superscript refers to which hypothesis, and the subscript refers to which time step.  The number \\(c\\) refers to the confidence.  You can only have up to three path hypotheses, so \\(k\\) can only be 0, 1, or 2.  You assign the confidence to the entire path, not to individual time steps.  This means there are a total of three confidences \\(c^k\\), and they are \\(c^0\\),\\( c^1\\), and \\(c^2\\).  Hope this helps.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1049111": "Does \\\\(c^k\\\\) in the evaluation metric equation mean \\\\(c_k\\\\) instead?",
    "1057632": "I believe in the evaluation metric, the superscript refers to which hypothesis, and the subscript refers to which time step.  The number \\\\(c\\\\) refers to the confidence.  You can only have up to three path hypotheses, so \\\\(k\\\\) can only be 0, 1, or 2.  You assign the confidence to the entire path, not to individual time steps.  This means there are a total of three confidences \\\\(c^k\\\\), and they are \\\\(c^0\\\\),\\\\( c^1\\\\), and \\\\(c^2\\\\).  Hope this helps."
  },
  "source": "meta"
}