{
  "id": 308355,
  "title": "MAP@k Evaluation Metric ",
  "url": "/competitions/h-and-m-personalized-fashion-recommendations/discussion/308355",
  "author_name": "Nikhil Gulwade",
  "post_date": "2022-02-18T09:02:16.850000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>MAP at k (Mean Average Precision at cutoff k):</p>\n<p>Precision and Recall metric does not account ordering while recommending whereas <br>\nPrecision at cutoff k is the precision measured on recommendations from rank 1 through k, for this exercise its 12.</p>\n<p>for ex: We've done 3 recommendations [0, 1, 1]. where 0 means incorrect recommendation and 1 means correct recommendation. <br>\nThen the precision at k = [0, 1/2, 2/3] and the average precision will be (1/5)*(0+1/2+2/3) = 0.38</p>\n<p>Larger the mean average precision, more correct will be the recommendations.</p>",
  "messages": [
    {
      "id": 1695616,
      "postDate": "2022-02-18T09:02:16.850Z",
      "content": "<p>MAP at k (Mean Average Precision at cutoff k):</p>\n<p>Precision and Recall metric does not account ordering while recommending whereas <br>\nPrecision at cutoff k is the precision measured on recommendations from rank 1 through k, for this exercise its 12.</p>\n<p>for ex: We've done 3 recommendations [0, 1, 1]. where 0 means incorrect recommendation and 1 means correct recommendation. <br>\nThen the precision at k = [0, 1/2, 2/3] and the average precision will be (1/5)*(0+1/2+2/3) = 0.38</p>\n<p>Larger the mean average precision, more correct will be the recommendations.</p>",
      "rawMarkdown": "MAP at k (Mean Average Precision at cutoff k):\n\nPrecision and Recall metric does not account ordering while recommending whereas \nPrecision at cutoff k is the precision measured on recommendations from rank 1 through k, for this exercise its 12.\n\nfor ex: We've done 3 recommendations [0, 1, 1]. where 0 means incorrect recommendation and 1 means correct recommendation. \nThen the precision at k = [0, 1/2, 2/3] and the average precision will be (1/5)*(0+1/2+2/3) = 0.38\n\nLarger the mean average precision, more correct will be the recommendations.",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1695616": "MAP at k (Mean Average Precision at cutoff k):\n\nPrecision and Recall metric does not account ordering while recommending whereas \nPrecision at cutoff k is the precision measured on recommendations from rank 1 through k, for this exercise its 12.\n\nfor ex: We've done 3 recommendations [0, 1, 1]. where 0 means incorrect recommendation and 1 means correct recommendation. \nThen the precision at k = [0, 1/2, 2/3] and the average precision will be (1/5)*(0+1/2+2/3) = 0.38\n\nLarger the mean average precision, more correct will be the recommendations."
  }
}