{
  "id": 372552,
  "title": "p(cart, click) = p(cart | click) p(click)",
  "url": "/competitions/otto-recommender-system/discussion/372552",
  "author_name": "toshi_k",
  "post_date": "2022-12-16T15:51:36.398000",
  "votes": 10,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Do you predict cart/order items from clicked items ?</p>\n<p>According to Bayes' theorem, joint distribution of cart and click can be modeled as below.</p>\n<blockquote>\n  <p>p(cart, click) = p(cart | click) p(click)</p>\n</blockquote>\n<p>p(cart | click) may be modeled easier than p(cart) or p(cart, click) directly.</p>",
  "messages": [
    {
      "id": 2067333,
      "postDate": "2022-12-16T15:51:36.400Z",
      "content": "<p>Do you predict cart/order items from clicked items ?</p>\n<p>According to Bayes' theorem, joint distribution of cart and click can be modeled as below.</p>\n<blockquote>\n  <p>p(cart, click) = p(cart | click) p(click)</p>\n</blockquote>\n<p>p(cart | click) may be modeled easier than p(cart) or p(cart, click) directly.</p>",
      "rawMarkdown": "Do you predict cart/order items from clicked items ?\n\nAccording to Bayes' theorem, joint distribution of cart and click can be modeled as below.\n>p(cart, click) = p(cart | click) p(click)\n\np(cart | click) may be modeled easier than p(cart) or p(cart, click) directly.",
      "votes": 9
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2067333": "Do you predict cart/order items from clicked items ?\n\nAccording to Bayes' theorem, joint distribution of cart and click can be modeled as below.\n>p(cart, click) = p(cart | click) p(click)\n\np(cart | click) may be modeled easier than p(cart) or p(cart, click) directly."
  }
}