{
  "id": 68945,
  "title": "Anyone used fast.ai library on this problem?",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/68945",
  "author_name": "",
  "post_date": "2018-10-18T18:05:02.710143400Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hi!\nI wonder if anyone tried to solve this competition using the fast.ai library?</p>",
  "messages": [
    {
      "id": "406137",
      "postDate": "10/18/2018 18:05:02",
      "content": "<p>Hi!\nI wonder if anyone tried to solve this competition using the fast.ai library?</p>",
      "rawMarkdown": "Hi!\nI wonder if anyone tried to solve this competition using the fast.ai library?",
      "votes": null
    },
    {
      "id": "406617",
      "postDate": "10/19/2018 14:31:06",
      "content": "<p>I did a very quick and shallow comparison between LGBM and fast ai.</p>\n\n<p>Same features, using one hot encoding for LGBM (only using top categories for high cardinality columns) vs. categorical embeddings with fast ai.</p>\n\n<p>fast ai 1.4533\nLGBM 1.4444</p>",
      "rawMarkdown": "I did a very quick and shallow comparison between LGBM and fast ai.\n\nSame features, using one hot encoding for LGBM (only using top categories for high cardinality columns) vs. categorical embeddings with fast ai.\n\nfast ai 1.4533\nLGBM 1.4444",
      "votes": null
    },
    {
      "id": "412175",
      "postDate": "10/29/2018 18:18:55",
      "content": "<p>hey can you make a kernel about the usage of fast.ai in this problem.</p>",
      "rawMarkdown": "hey can you make a kernel about the usage of fast.ai in this problem.",
      "votes": null
    },
    {
      "id": "521282",
      "postDate": "04/22/2019 17:35:25",
      "content": "<p>I am not sure if the FastAI library has support for R</p>",
      "rawMarkdown": "I am not sure if the FastAI library has support for R",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 406617,
      "author_name": "sdoria",
      "author_url": "",
      "post_date": "10/19/2018 14:31:06",
      "content": "<p>I did a very quick and shallow comparison between LGBM and fast ai.</p>\n\n<p>Same features, using one hot encoding for LGBM (only using top categories for high cardinality columns) vs. categorical embeddings with fast ai.</p>\n\n<p>fast ai 1.4533\nLGBM 1.4444</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 412175,
      "author_name": "vishwasgpai",
      "author_url": "",
      "post_date": "10/29/2018 18:18:55",
      "content": "<p>hey can you make a kernel about the usage of fast.ai in this problem.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 521282,
      "author_name": "adeebabbas",
      "author_url": "",
      "post_date": "04/22/2019 17:35:25",
      "content": "<p>I am not sure if the FastAI library has support for R</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "406137": "Hi!\nI wonder if anyone tried to solve this competition using the fast.ai library?",
    "406617": "I did a very quick and shallow comparison between LGBM and fast ai.\n\nSame features, using one hot encoding for LGBM (only using top categories for high cardinality columns) vs. categorical embeddings with fast ai.\n\nfast ai 1.4533\nLGBM 1.4444",
    "412175": "hey can you make a kernel about the usage of fast.ai in this problem.",
    "521282": "I am not sure if the FastAI library has support for R"
  },
  "source": "meta"
}