{
  "id": 7869,
  "title": "newbie's question",
  "url": "/competitions/acquire-valued-shoppers-challenge/discussion/7869",
  "author_name": "",
  "post_date": "2014-04-25T14:50:34.890Z",
  "votes": null,
  "comment_count": 6,
  "views": 1918,
  "content": "<p>I submited the &quot;sampleSubmission.csv&quot; file,the score is 0.5 .Is 0.5 the auc?how is it&nbsp; computed according to the file i submited?</p>",
  "messages": [
    {
      "id": "43006",
      "postDate": "04/25/2014 14:50:34",
      "content": "<p>I submited the &quot;sampleSubmission.csv&quot; file,the score is 0.5 .Is 0.5 the auc?how is it&nbsp; computed according to the file i submited?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43007",
      "postDate": "04/25/2014 15:15:49",
      "content": "<p>Sample submission contains all zeros. The 0.5 score on the leaderboard is Area under the ROC curve (AUC). If you want to learn in detail how AUC is calculated:&nbsp;http://en.wikipedia.org/wiki/Receiver_operating_characteristic</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43012",
      "postDate": "04/25/2014 16:28:24",
      "content": "<p>Thanks.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43077",
      "postDate": "04/26/2014 12:56:25",
      "content": "<p>I am learning data mining through R.&nbsp;</p>\n<p>Can you suggest how to start with this dataset.&nbsp;</p>\n<p>A stepwise approach may be.&nbsp;</p>\n\n<p>Thanks</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43078",
      "postDate": "04/26/2014 14:03:55",
      "content": "<p>I imagine people will hang onto some secrets until the competition ends. &nbsp;If you're starting with data mining you may want to look at a finished competition where people have revealed what they did (and in recent cases, code too). &nbsp;There's an older&nbsp;predicting shopping challenge at&nbsp;<a href=\"http://www.kaggle.com/c/dunnhumbychallenge\">http://www.kaggle.com/c/dunnhumbychallenge</a>.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43079",
      "postDate": "04/26/2014 14:22:28",
      "content": "<p>Mlearn thanks.&nbsp;</p>\n<p>In general what are the steps you choose to start with-&nbsp;</p>\n<p>data cleaning</p>\n<p>reduction&nbsp;</p>\n<p>model( i find it difficult to understand which one will be more suitable, do we have any source to check for such kind of problem? )</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43095",
      "postDate": "04/26/2014 18:59:53",
      "content": "<p>Any constant value submission will result in an AUC of 0.5.</p>\n\n<p>One starting point would be to replicate the benchmarks. &nbsp;For each offer in the test set, lookup in the customer previously bought that brand, that company, or that category.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 43007,
      "author_name": "abhishek",
      "author_url": "",
      "post_date": "04/25/2014 15:15:49",
      "content": "<p>Sample submission contains all zeros. The 0.5 score on the leaderboard is Area under the ROC curve (AUC). If you want to learn in detail how AUC is calculated:&nbsp;http://en.wikipedia.org/wiki/Receiver_operating_characteristic</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43012,
      "author_name": "henryfung",
      "author_url": "",
      "post_date": "04/25/2014 16:28:24",
      "content": "<p>Thanks.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43077,
      "author_name": "chandnijpurohit",
      "author_url": "",
      "post_date": "04/26/2014 12:56:25",
      "content": "<p>I am learning data mining through R.&nbsp;</p>\n<p>Can you suggest how to start with this dataset.&nbsp;</p>\n<p>A stepwise approach may be.&nbsp;</p>\n\n<p>Thanks</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43078,
      "author_name": "mlearn",
      "author_url": "",
      "post_date": "04/26/2014 14:03:55",
      "content": "<p>I imagine people will hang onto some secrets until the competition ends. &nbsp;If you're starting with data mining you may want to look at a finished competition where people have revealed what they did (and in recent cases, code too). &nbsp;There's an older&nbsp;predicting shopping challenge at&nbsp;<a href=\"http://www.kaggle.com/c/dunnhumbychallenge\">http://www.kaggle.com/c/dunnhumbychallenge</a>.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43079,
      "author_name": "chandnijpurohit",
      "author_url": "",
      "post_date": "04/26/2014 14:22:28",
      "content": "<p>Mlearn thanks.&nbsp;</p>\n<p>In general what are the steps you choose to start with-&nbsp;</p>\n<p>data cleaning</p>\n<p>reduction&nbsp;</p>\n<p>model( i find it difficult to understand which one will be more suitable, do we have any source to check for such kind of problem? )</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43095,
      "author_name": "zachmayer",
      "author_url": "",
      "post_date": "04/26/2014 18:59:53",
      "content": "<p>Any constant value submission will result in an AUC of 0.5.</p>\n\n<p>One starting point would be to replicate the benchmarks. &nbsp;For each offer in the test set, lookup in the customer previously bought that brand, that company, or that category.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "43006": "",
    "43007": "",
    "43012": "",
    "43077": "",
    "43078": "",
    "43079": "",
    "43095": ""
  },
  "source": "meta"
}