{
  "id": 8141,
  "title": "How far in the future are we predicting?",
  "url": "/competitions/acquire-valued-shoppers-challenge/discussion/8141",
  "author_name": "",
  "post_date": "2014-05-14T02:09:13.367Z",
  "votes": null,
  "comment_count": 6,
  "views": 1954,
  "content": "<p>In the problem description it is mentioned that we should&nbsp;predict&nbsp;the probability of customer&nbsp;return (during or after the promotional period) to purchase the same item again.</p>\n<p>How many days/months is this period from the&nbsp;offer date or the last transaction?</p>",
  "messages": [
    {
      "id": "44493",
      "postDate": "05/14/2014 02:09:13",
      "content": "<p>In the problem description it is mentioned that we should&nbsp;predict&nbsp;the probability of customer&nbsp;return (during or after the promotional period) to purchase the same item again.</p>\n<p>How many days/months is this period from the&nbsp;offer date or the last transaction?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "44524",
      "postDate": "05/14/2014 13:07:24",
      "content": "<p>I think that it's calculated by some fixed period transactions data. Like they sliced it till September '13 (or other month)</p>\n<p>If it's so it is possible to reverse engineering&nbsp;this counting repeats from customer's that have offers and plot graph &quot;offer date (roughed to weeks for example) - mean count of repeats for&nbsp;users with offer in this date&quot;. If it's linear you can extrapolate it until intersection with &quot;0&quot; and round to month start. If my assumption is true. If it's roughly constant you can't get much value from this I&nbsp;think, but in this case you also can't extract much value from this move.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "44649",
      "postDate": "05/16/2014 07:23:13",
      "content": "<p>Hi, I think they are using historical data to do the back testing. &nbsp;With the customer rank that you provide, they will cross verify with their actual records (which is not published) and use the matching rate as the judgement. &nbsp;So, the number of predicting months should not be the parameters in the equation to calculate user score, i.e.&nbsp;number of predicting months does not matter.&nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "44650",
      "postDate": "05/16/2014 07:51:47",
      "content": "<p>Can you clarify that. I don't understand the point</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "44654",
      "postDate": "05/16/2014 08:59:18",
      "content": "<p>I think that the prediction period is from the offerdate until the date when the data was extracted from the database.</p>\n<p>The probability of return depends on the length of this period, and I think that this information should be provided.</p>\n<p>Am I missing something here?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "44661",
      "postDate": "05/16/2014 10:59:21",
      "content": "<p>I think the implicit&nbsp;rule for kaggle competitions : admins&nbsp;can clarify but generally you can't get more data, so if you want to have this info, you should extract it from data.<br>But it will be very helpful&nbsp;to get comment from admin about&nbsp;your question and my opinion :)</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "47328",
      "postDate": "05/30/2014 11:24:30",
      "content": "<p>Thankyou..</p>\n<p>It would be great if admin can clarify this.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 44524,
      "author_name": "dremovd",
      "author_url": "",
      "post_date": "05/14/2014 13:07:24",
      "content": "<p>I think that it's calculated by some fixed period transactions data. Like they sliced it till September '13 (or other month)</p>\n<p>If it's so it is possible to reverse engineering&nbsp;this counting repeats from customer's that have offers and plot graph &quot;offer date (roughed to weeks for example) - mean count of repeats for&nbsp;users with offer in this date&quot;. If it's linear you can extrapolate it until intersection with &quot;0&quot; and round to month start. If my assumption is true. If it's roughly constant you can't get much value from this I&nbsp;think, but in this case you also can't extract much value from this move.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 44649,
      "author_name": "jimmyzhang",
      "author_url": "",
      "post_date": "05/16/2014 07:23:13",
      "content": "<p>Hi, I think they are using historical data to do the back testing. &nbsp;With the customer rank that you provide, they will cross verify with their actual records (which is not published) and use the matching rate as the judgement. &nbsp;So, the number of predicting months should not be the parameters in the equation to calculate user score, i.e.&nbsp;number of predicting months does not matter.&nbsp;</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 44650,
      "author_name": "dremovd",
      "author_url": "",
      "post_date": "05/16/2014 07:51:47",
      "content": "<p>Can you clarify that. I don't understand the point</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 44654,
      "author_name": "adilmouja",
      "author_url": "",
      "post_date": "05/16/2014 08:59:18",
      "content": "<p>I think that the prediction period is from the offerdate until the date when the data was extracted from the database.</p>\n<p>The probability of return depends on the length of this period, and I think that this information should be provided.</p>\n<p>Am I missing something here?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 44661,
      "author_name": "dremovd",
      "author_url": "",
      "post_date": "05/16/2014 10:59:21",
      "content": "<p>I think the implicit&nbsp;rule for kaggle competitions : admins&nbsp;can clarify but generally you can't get more data, so if you want to have this info, you should extract it from data.<br>But it will be very helpful&nbsp;to get comment from admin about&nbsp;your question and my opinion :)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 47328,
      "author_name": "saikumarallaka",
      "author_url": "",
      "post_date": "05/30/2014 11:24:30",
      "content": "<p>Thankyou..</p>\n<p>It would be great if admin can clarify this.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "44493": "",
    "44524": "",
    "44649": "",
    "44650": "",
    "44654": "",
    "44661": "",
    "47328": ""
  },
  "source": "meta"
}