{
  "id": 82069,
  "title": "Congratulations to all winners! For me its a costly lession learned!!",
  "url": "/competitions/elo-merchant-category-recommendation/discussion/82069",
  "author_name": "",
  "post_date": "2019-02-27T04:11:31.955684800Z",
  "votes": 10,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Congratulations to all winners. \nI made a costly mistake by not considering month old submissions as a final one. It could have fetched me a silver probably (3.611). The lesson is to have a  strong local CV measurement/tracking mechanism and not get carried away by the latest submissions. I went on overfitting to public LB while lost on private LB.\nHope I will do better next time. \nCongratulations again!</p>",
  "messages": [
    {
      "id": "479315",
      "postDate": "02/27/2019 04:11:31",
      "content": "<p>Congratulations to all winners. \nI made a costly mistake by not considering month old submissions as a final one. It could have fetched me a silver probably (3.611). The lesson is to have a  strong local CV measurement/tracking mechanism and not get carried away by the latest submissions. I went on overfitting to public LB while lost on private LB.\nHope I will do better next time. \nCongratulations again!</p>",
      "rawMarkdown": "Congratulations to all winners. \nI made a costly mistake by not considering month old submissions as a final one. It could have fetched me a silver probably (3.611). The lesson is to have a  strong local CV measurement/tracking mechanism and not get carried away by the latest submissions. I went on overfitting to public LB while lost on private LB.\nHope I will do better next time. \nCongratulations again!",
      "votes": null
    },
    {
      "id": "479432",
      "postDate": "02/27/2019 06:09:31",
      "content": "<p>This competition is easily to be overfitted, and you have done your best.</p>",
      "rawMarkdown": "This competition is easily to be overfitted, and you have done your best.",
      "votes": null
    },
    {
      "id": "479448",
      "postDate": "02/27/2019 06:34:43",
      "content": "<p>Agree!</p>",
      "rawMarkdown": "Agree!",
      "votes": null
    },
    {
      "id": "479684",
      "postDate": "02/27/2019 10:40:28",
      "content": "<p>I made the same mistake on previous Santander competition, learned the hard way, \nIt allowed me not to make the same mistake again =)\nIt feels awful to see that a significantly better submission was not selected.</p>\n\n<p>Here's to lessons learned,\nGood job!!</p>",
      "rawMarkdown": "I made the same mistake on previous Santander competition, learned the hard way, \nIt allowed me not to make the same mistake again =)\nIt feels awful to see that a significantly better submission was not selected.\n\nHere's to lessons learned,\nGood job!!",
      "votes": null
    },
    {
      "id": "479804",
      "postDate": "02/27/2019 12:46:15",
      "content": "<p>Don't feel bad <a href=\"/prasunmishra\">@prasunmishra</a> and put it down to lesson learned. I have experienced exactly the same thing in the past and only escaped that fate in this competition due to the progress I made in the last 3 days. This data is an interesting and difficult one to deal with. In that sense it is much closer to real life ML/ Data Science problem from the work place.</p>",
      "rawMarkdown": "Don't feel bad @prasunmishra and put it down to lesson learned. I have experienced exactly the same thing in the past and only escaped that fate in this competition due to the progress I made in the last 3 days. This data is an interesting and difficult one to deal with. In that sense it is much closer to real life ML/ Data Science problem from the work place.",
      "votes": null
    },
    {
      "id": "480082",
      "postDate": "02/27/2019 18:57:02",
      "content": "<p>Thanks @YaGana!</p>",
      "rawMarkdown": "Thanks @YaGana!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 479432,
      "author_name": "liewmanchoi",
      "author_url": "",
      "post_date": "02/27/2019 06:09:31",
      "content": "<p>This competition is easily to be overfitted, and you have done your best.</p>",
      "votes": null,
      "replies": [
        {
          "id": 479448,
          "author_name": "prasunmishra",
          "author_url": "",
          "post_date": "02/27/2019 06:34:43",
          "content": "<p>Agree!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 479684,
      "author_name": "tpthegreat",
      "author_url": "",
      "post_date": "02/27/2019 10:40:28",
      "content": "<p>I made the same mistake on previous Santander competition, learned the hard way, \nIt allowed me not to make the same mistake again =)\nIt feels awful to see that a significantly better submission was not selected.</p>\n\n<p>Here's to lessons learned,\nGood job!!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 479804,
      "author_name": "sheriytm",
      "author_url": "",
      "post_date": "02/27/2019 12:46:15",
      "content": "<p>Don't feel bad <a href=\"/prasunmishra\">@prasunmishra</a> and put it down to lesson learned. I have experienced exactly the same thing in the past and only escaped that fate in this competition due to the progress I made in the last 3 days. This data is an interesting and difficult one to deal with. In that sense it is much closer to real life ML/ Data Science problem from the work place.</p>",
      "votes": null,
      "replies": [
        {
          "id": 480082,
          "author_name": "prasunmishra",
          "author_url": "",
          "post_date": "02/27/2019 18:57:02",
          "content": "<p>Thanks @YaGana!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "479315": "Congratulations to all winners. \nI made a costly mistake by not considering month old submissions as a final one. It could have fetched me a silver probably (3.611). The lesson is to have a  strong local CV measurement/tracking mechanism and not get carried away by the latest submissions. I went on overfitting to public LB while lost on private LB.\nHope I will do better next time. \nCongratulations again!",
    "479432": "This competition is easily to be overfitted, and you have done your best.",
    "479448": "Agree!",
    "479684": "I made the same mistake on previous Santander competition, learned the hard way, \nIt allowed me not to make the same mistake again =)\nIt feels awful to see that a significantly better submission was not selected.\n\nHere's to lessons learned,\nGood job!!",
    "479804": "Don't feel bad @prasunmishra and put it down to lesson learned. I have experienced exactly the same thing in the past and only escaped that fate in this competition due to the progress I made in the last 3 days. This data is an interesting and difficult one to deal with. In that sense it is much closer to real life ML/ Data Science problem from the work place.",
    "480082": "Thanks @YaGana!"
  },
  "source": "meta"
}