{
  "id": 114896,
  "title": "Side Competition",
  "url": "/competitions/ashrae-energy-prediction/discussion/114896",
  "author_name": "",
  "post_date": "2019-10-29T23:03:22.884251500Z",
  "votes": 3,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Can you predict how soon the blending will start?</p>",
  "messages": [
    {
      "id": "661052",
      "postDate": "10/29/2019 23:03:22",
      "content": "<p>Can you predict how soon the blending will start?</p>",
      "rawMarkdown": "Can you predict how soon the blending will start?",
      "votes": null
    },
    {
      "id": "661189",
      "postDate": "10/30/2019 02:00:27",
      "content": "<p>Participated in large number of competitions in last years.  My casual observation is that blending starts when a space of 10 days occurs without someone sharing a new low score kernel.</p>",
      "rawMarkdown": "Participated in large number of competitions in last years.  My casual observation is that blending starts when a space of 10 days occurs without someone sharing a new low score kernel.",
      "votes": null
    },
    {
      "id": "661416",
      "postDate": "10/30/2019 09:19:10",
      "content": "<p>I reckon it has already started. ;)</p>",
      "rawMarkdown": "I reckon it has already started. ;)",
      "votes": null
    },
    {
      "id": "661417",
      "postDate": "10/30/2019 09:19:28",
      "content": "<p>As far as i know, people are already doing blending. We have seen NN, catboost, xgboost, lightgbm models doing pretty good themselves. So it is expected to blend the best models among them.</p>",
      "rawMarkdown": "As far as i know, people are already doing blending. We have seen NN, catboost, xgboost, lightgbm models doing pretty good themselves. So it is expected to blend the best models among them.",
      "votes": null
    },
    {
      "id": "667188",
      "postDate": "11/06/2019 23:38:25",
      "content": "<p>We have a winner! <a href=\"https://www.kaggle.com/hmendonca/4-ashrae-blended/output\">https://www.kaggle.com/hmendonca/4-ashrae-blended/output</a></p>",
      "rawMarkdown": "We have a winner! https://www.kaggle.com/hmendonca/4-ashrae-blended/output",
      "votes": null
    },
    {
      "id": "667251",
      "postDate": "11/07/2019 02:38:05",
      "content": "<p>What is blending? </p>",
      "rawMarkdown": "What is blending?",
      "votes": null
    },
    {
      "id": "667357",
      "postDate": "11/07/2019 06:30:22",
      "content": "<p><a href=\"https://www.kaggle.com/hmendonca/4-ashrae-blended\">https://www.kaggle.com/hmendonca/4-ashrae-blended</a></p>",
      "rawMarkdown": "https://www.kaggle.com/hmendonca/4-ashrae-blended",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 661189,
      "author_name": "pcjimmmy",
      "author_url": "",
      "post_date": "10/30/2019 02:00:27",
      "content": "<p>Participated in large number of competitions in last years.  My casual observation is that blending starts when a space of 10 days occurs without someone sharing a new low score kernel.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 661416,
      "author_name": "frjo0083",
      "author_url": "",
      "post_date": "10/30/2019 09:19:10",
      "content": "<p>I reckon it has already started. ;)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 661417,
      "author_name": "khairulislam",
      "author_url": "",
      "post_date": "10/30/2019 09:19:28",
      "content": "<p>As far as i know, people are already doing blending. We have seen NN, catboost, xgboost, lightgbm models doing pretty good themselves. So it is expected to blend the best models among them.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 667188,
      "author_name": "gideonvos",
      "author_url": "",
      "post_date": "11/06/2019 23:38:25",
      "content": "<p>We have a winner! <a href=\"https://www.kaggle.com/hmendonca/4-ashrae-blended/output\">https://www.kaggle.com/hmendonca/4-ashrae-blended/output</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 667251,
      "author_name": "diginfo",
      "author_url": "",
      "post_date": "11/07/2019 02:38:05",
      "content": "<p>What is blending? </p>",
      "votes": null,
      "replies": [
        {
          "id": 667357,
          "author_name": "gideonvos",
          "author_url": "",
          "post_date": "11/07/2019 06:30:22",
          "content": "<p><a href=\"https://www.kaggle.com/hmendonca/4-ashrae-blended\">https://www.kaggle.com/hmendonca/4-ashrae-blended</a></p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "661052": "Can you predict how soon the blending will start?",
    "661189": "Participated in large number of competitions in last years.  My casual observation is that blending starts when a space of 10 days occurs without someone sharing a new low score kernel.",
    "661416": "I reckon it has already started. ;)",
    "661417": "As far as i know, people are already doing blending. We have seen NN, catboost, xgboost, lightgbm models doing pretty good themselves. So it is expected to blend the best models among them.",
    "667188": "We have a winner! https://www.kaggle.com/hmendonca/4-ashrae-blended/output",
    "667251": "What is blending?",
    "667357": "https://www.kaggle.com/hmendonca/4-ashrae-blended"
  },
  "source": "meta"
}