{
  "id": 1310,
  "title": "Usage of Data File",
  "url": "/competitions/benchmark-bond-trade-price-challenge/discussion/1310",
  "author_name": "",
  "post_date": "2012-01-28T16:57:02.753Z",
  "votes": null,
  "comment_count": 2,
  "views": 2282,
  "content": "<p>Hi, All&nbsp;</p>\r\n<p><span style=\"font-size:small\">Sorry it may seems stupid . &nbsp;It's my first time here ,not seeing any explaination about the data file&nbsp;</span></p>\r\n<p><span class=\"x_Apple-style-span\" style=\"line-height:normal; font-size:small\">test // &nbsp;is the history bond trading data ? we base this data to build up model ?</span><span class=\"x_Apple-style-span\" style=\"line-height:normal; font-size:small\">&nbsp;is this also\r\n history bond trading data ,but used to evaluate the accracy of models we build ?</span></p>\r\n<p><span style=\"font-size:small\"><span style=\"font-size:small\"><span class=\"x_x_Apple-style-span\" style=\"line-height:normal\">train // is this bond trading data used for Machine learning / AI training ?</span></span></span></p>\r\n<p><span style=\"font-size:small\">for me they are all history data with no difference ..is that right ? Thank you in advance for explaining this .</span></p>",
  "messages": [
    {
      "id": "8310",
      "postDate": "01/28/2012 16:57:02",
      "content": "<p>Hi, All&nbsp;</p>\r\n<p><span style=\"font-size:small\">Sorry it may seems stupid . &nbsp;It's my first time here ,not seeing any explaination about the data file&nbsp;</span></p>\r\n<p><span class=\"x_Apple-style-span\" style=\"line-height:normal; font-size:small\">test // &nbsp;is the history bond trading data ? we base this data to build up model ?</span><span class=\"x_Apple-style-span\" style=\"line-height:normal; font-size:small\">&nbsp;is this also\r\n history bond trading data ,but used to evaluate the accracy of models we build ?</span></p>\r\n<p><span style=\"font-size:small\"><span style=\"font-size:small\"><span class=\"x_x_Apple-style-span\" style=\"line-height:normal\">train // is this bond trading data used for Machine learning / AI training ?</span></span></span></p>\r\n<p><span style=\"font-size:small\">for me they are all history data with no difference ..is that right ? Thank you in advance for explaining this .</span></p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "8322",
      "postDate": "01/28/2012 22:47:20",
      "content": "<p>In the train data you have the &quot;answer&quot; which is trade_price. This is where you train your models. The test data is what you run your models against when you want to submit a solution.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "8346",
      "postDate": "01/30/2012 14:22:51",
      "content": "<p>What you need to do is use the train.csv file to come up with a model. The model needs to output/predict the trade_price value, all other fields will form your model inputs. Once you are convinced you have a winning model, you have to use the data from the\r\n test.csv file (no trade_price in here) and predict the trade_price. Your predictions need to be recorded as shown in the random_forest_sample_submission.csv example to be submitted successfully.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 8322,
      "author_name": "danglaser",
      "author_url": "",
      "post_date": "01/28/2012 22:47:20",
      "content": "<p>In the train data you have the &quot;answer&quot; which is trade_price. This is where you train your models. The test data is what you run your models against when you want to submit a solution.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 8346,
      "author_name": "evangelos2",
      "author_url": "",
      "post_date": "01/30/2012 14:22:51",
      "content": "<p>What you need to do is use the train.csv file to come up with a model. The model needs to output/predict the trade_price value, all other fields will form your model inputs. Once you are convinced you have a winning model, you have to use the data from the\r\n test.csv file (no trade_price in here) and predict the trade_price. Your predictions need to be recorded as shown in the random_forest_sample_submission.csv example to be submitted successfully.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "8310": "",
    "8322": "",
    "8346": ""
  },
  "source": "meta"
}