{
  "id": 71092,
  "title": "How to predict further ？",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/71092",
  "author_name": "",
  "post_date": "2018-11-10T02:25:05.488113500Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I can’t understand how to predict further？what is the input and output of the model？what is the data organization strategy ？</p>",
  "messages": [
    {
      "id": "418516",
      "postDate": "11/10/2018 02:25:05",
      "content": "<p>I can’t understand how to predict further？what is the input and output of the model？what is the data organization strategy ？</p>",
      "rawMarkdown": "I can’t understand how to predict further？what is the input and output of the model？what is the data organization strategy ？",
      "votes": null
    },
    {
      "id": "419374",
      "postDate": "11/11/2018 20:32:21",
      "content": "<p>Some naive ideas:\n- input: customer's historical events +  user features\n- output: Customer's future revenue </p>",
      "rawMarkdown": "Some naive ideas:\n- input: customer's historical events +  user features\n- output: Customer's future revenue",
      "votes": null
    },
    {
      "id": "419652",
      "postDate": "11/12/2018 11:07:41",
      "content": "<p>So do we just make up some fake columns then make predictions with those?</p>",
      "rawMarkdown": "So do we just make up some fake columns then make predictions with those?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 419374,
      "author_name": "fangzheng13",
      "author_url": "",
      "post_date": "11/11/2018 20:32:21",
      "content": "<p>Some naive ideas:\n- input: customer's historical events +  user features\n- output: Customer's future revenue </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 419652,
      "author_name": "icemekaveli",
      "author_url": "",
      "post_date": "11/12/2018 11:07:41",
      "content": "<p>So do we just make up some fake columns then make predictions with those?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "418516": "I can’t understand how to predict further？what is the input and output of the model？what is the data organization strategy ？",
    "419374": "Some naive ideas:\n- input: customer's historical events +  user features\n- output: Customer's future revenue",
    "419652": "So do we just make up some fake columns then make predictions with those?"
  },
  "source": "meta"
}