{
  "id": 327696,
  "title": "What I found !",
  "url": "/competitions/amex-default-prediction/discussion/327696",
  "author_name": "",
  "post_date": "2022-05-28T15:16:27.636174400Z",
  "votes": 12,
  "comment_count": 3,
  "views": 0,
  "content": "<p>In this discussion, I will post what I found in EDA.<br>\nI hope to update this little by little.</p>\n<p>If I am wrong or you find other insights, please comment!!</p>",
  "messages": [
    {
      "id": "1804150",
      "postDate": "05/28/2022 15:16:27",
      "content": "<p>In this discussion, I will post what I found in EDA.<br>\nI hope to update this little by little.</p>\n<p>If I am wrong or you find other insights, please comment!!</p>",
      "rawMarkdown": "In this discussion, I will post what I found in EDA.\nI hope to update this little by little.\n\nIf I am wrong or you find other insights, please comment!!",
      "votes": null
    },
    {
      "id": "1804153",
      "postDate": "05/28/2022 15:18:42",
      "content": "<p>There are many more like these. keep on digging and enjoy :)</p>",
      "rawMarkdown": "There are many more like these. keep on digging and enjoy :)",
      "votes": null
    },
    {
      "id": "1804471",
      "postDate": "05/29/2022 03:24:18",
      "content": "<p>I found 'B_31' and 'D_87' have a few unique values.<br>\nIn train and test, B_31 has 0, 1 and Nan and D_87 has 1.0 and Nan.</p>\n<p>But these columns are not listed as categorical feature in the data description page.<br>\nAre these columns categorical feature?</p>",
      "rawMarkdown": "I found 'B_31' and 'D_87' have a few unique values.\nIn train and test, B_31 has 0, 1 and Nan and D_87 has 1.0 and Nan.\n\nBut these columns are not listed as categorical feature in the data description page.\nAre these columns categorical feature?",
      "votes": null
    },
    {
      "id": "1804473",
      "postDate": "05/29/2022 03:30:06",
      "content": "<p>Many customer have all monthly records from 2017/3 ~ 2018/3.<br>\nBut a part of customer have few records.<br>\nThese customers may sign up in the middle of the term or have months of non-use.</p>\n<pre><code>1     0.335742\n2     0.318465\n3     0.358602\n4     0.416221\n5     0.392635\n6     0.387670\n7     0.418430\n8     0.447300\n9     0.450164\n10    0.462282\n11    0.446737\n12    0.389344\n13    0.231788\n</code></pre>\n<p>The above list shows the different ratios of targets for different numbers of records customers have. This could be used for useful features.</p>",
      "rawMarkdown": "Many customer have all monthly records from 2017/3 ~ 2018/3.\nBut a part of customer have few records.\nThese customers may sign up in the middle of the term or have months of non-use.\n\n```\n1     0.335742\n2     0.318465\n3     0.358602\n4     0.416221\n5     0.392635\n6     0.387670\n7     0.418430\n8     0.447300\n9     0.450164\n10    0.462282\n11    0.446737\n12    0.389344\n13    0.231788\n```\n\nThe above list shows the different ratios of targets for different numbers of records customers have. This could be used for useful features.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1804153,
      "author_name": "raddar",
      "author_url": "",
      "post_date": "05/28/2022 15:18:42",
      "content": "<p>There are many more like these. keep on digging and enjoy :)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1804471,
      "author_name": "ryotak12",
      "author_url": "",
      "post_date": "05/29/2022 03:24:18",
      "content": "<p>I found 'B_31' and 'D_87' have a few unique values.<br>\nIn train and test, B_31 has 0, 1 and Nan and D_87 has 1.0 and Nan.</p>\n<p>But these columns are not listed as categorical feature in the data description page.<br>\nAre these columns categorical feature?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1804473,
      "author_name": "ryotak12",
      "author_url": "",
      "post_date": "05/29/2022 03:30:06",
      "content": "<p>Many customer have all monthly records from 2017/3 ~ 2018/3.<br>\nBut a part of customer have few records.<br>\nThese customers may sign up in the middle of the term or have months of non-use.</p>\n<pre><code>1     0.335742\n2     0.318465\n3     0.358602\n4     0.416221\n5     0.392635\n6     0.387670\n7     0.418430\n8     0.447300\n9     0.450164\n10    0.462282\n11    0.446737\n12    0.389344\n13    0.231788\n</code></pre>\n<p>The above list shows the different ratios of targets for different numbers of records customers have. This could be used for useful features.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1804150": "In this discussion, I will post what I found in EDA.\nI hope to update this little by little.\n\nIf I am wrong or you find other insights, please comment!!",
    "1804153": "There are many more like these. keep on digging and enjoy :)",
    "1804471": "I found 'B_31' and 'D_87' have a few unique values.\nIn train and test, B_31 has 0, 1 and Nan and D_87 has 1.0 and Nan.\n\nBut these columns are not listed as categorical feature in the data description page.\nAre these columns categorical feature?",
    "1804473": "Many customer have all monthly records from 2017/3 ~ 2018/3.\nBut a part of customer have few records.\nThese customers may sign up in the middle of the term or have months of non-use.\n\n```\n1     0.335742\n2     0.318465\n3     0.358602\n4     0.416221\n5     0.392635\n6     0.387670\n7     0.418430\n8     0.447300\n9     0.450164\n10    0.462282\n11    0.446737\n12    0.389344\n13    0.231788\n```\n\nThe above list shows the different ratios of targets for different numbers of records customers have. This could be used for useful features."
  },
  "source": "meta"
}