{
  "id": 371476,
  "title": "Fast Recall@20",
  "url": "/competitions/otto-recommender-system/discussion/371476",
  "author_name": "aldparis",
  "post_date": "2022-12-10T11:16:08.506000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi, <br>\nin the version 4 (and 5) of my previous <a href=\"https://www.kaggle.com/code/adaubas/otto-fast-handcrafted-model-recall-20\" target=\"_blank\">notebook</a>, I tried another way to compute Recall@20 : instead of doing intersection between candidates and ground truth on each row, I count when there are no difference between candidates columns and ground truth columns, to take advantage of the numpy's power.<br>\nIt could be usefull for later Features Permutation Importance in XGBRanker validation…</p>",
  "messages": [
    {
      "id": 2060747,
      "postDate": "2022-12-10T11:16:08.507Z",
      "content": "<p>Hi, <br>\nin the version 4 (and 5) of my previous <a href=\"https://www.kaggle.com/code/adaubas/otto-fast-handcrafted-model-recall-20\" target=\"_blank\">notebook</a>, I tried another way to compute Recall@20 : instead of doing intersection between candidates and ground truth on each row, I count when there are no difference between candidates columns and ground truth columns, to take advantage of the numpy's power.<br>\nIt could be usefull for later Features Permutation Importance in XGBRanker validation…</p>",
      "rawMarkdown": "Hi, \nin the version 4 (and 5) of my previous [notebook](https://www.kaggle.com/code/adaubas/otto-fast-handcrafted-model-recall-20), I tried another way to compute Recall@20 : instead of doing intersection between candidates and ground truth on each row, I count when there are no difference between candidates columns and ground truth columns, to take advantage of the numpy's power.\nIt could be usefull for later Features Permutation Importance in XGBRanker validation...",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2060747": "Hi, \nin the version 4 (and 5) of my previous [notebook](https://www.kaggle.com/code/adaubas/otto-fast-handcrafted-model-recall-20), I tried another way to compute Recall@20 : instead of doing intersection between candidates and ground truth on each row, I count when there are no difference between candidates columns and ground truth columns, to take advantage of the numpy's power.\nIt could be usefull for later Features Permutation Importance in XGBRanker validation..."
  }
}