{
  "id": 1837,
  "title": "Did anybody use the structure of weights (or time_last1)",
  "url": "/competitions/benchmark-bond-trade-price-challenge/discussion/1837",
  "author_name": "",
  "post_date": "2012-05-01T08:04:22.697Z",
  "votes": null,
  "comment_count": 2,
  "views": 3669,
  "content": "<p>Do a histogram of the weights and you'll see what I mean.&nbsp; the weights have a strong multimodal distribution, separated in sets corresponding to a period of about 26 hours (not 24...).</p>\r\n<p>Within each set of weights the median deviation between curve based price and actual price goes up linearly with time... I added that in my model mix and gained a tiny bit, but i was wondering what;s the phenomenon under that.</p>\r\n<p>My comp is busy right now but Ill make and post graphs later.&nbsp; If you have octave/matlab you can use the attached code to regen the graphs, the variables are self-explanatory...</p>\r\n<p>&nbsp;</p>\r\n<p>% train_data is the training set, Ntrain X 61</p>\r\n<p>% each graph plots a &quot;blob&quot; that is the all the data in the graph comes from trades within one of the modes of the multimodal distribution of delays between THE trade (indexes 1-11) and trade-1 (indexes 12-16 in data line).</p>\r\n<p>% each graph cuts that mode in 10 finer slices of time and plots the mean or median absolute error</p>\r\n<p>% the first 5 blobs, which comprise the vast majority of weight, show a strong relationship between time in the blob and abs(price-curve_based)</p>\r\n<p>&nbsp;</p>\r\n<p>% the last two graphs put together all blobs to illustrate the fact that the relationship gets a reset every blob.</p>\r\n<p>&nbsp;</p>\r\n<p>&nbsp;</p>",
  "messages": [
    {
      "id": "10710",
      "postDate": "05/01/2012 08:04:22",
      "content": "<p>Do a histogram of the weights and you'll see what I mean.&nbsp; the weights have a strong multimodal distribution, separated in sets corresponding to a period of about 26 hours (not 24...).</p>\r\n<p>Within each set of weights the median deviation between curve based price and actual price goes up linearly with time... I added that in my model mix and gained a tiny bit, but i was wondering what;s the phenomenon under that.</p>\r\n<p>My comp is busy right now but Ill make and post graphs later.&nbsp; If you have octave/matlab you can use the attached code to regen the graphs, the variables are self-explanatory...</p>\r\n<p>&nbsp;</p>\r\n<p>% train_data is the training set, Ntrain X 61</p>\r\n<p>% each graph plots a &quot;blob&quot; that is the all the data in the graph comes from trades within one of the modes of the multimodal distribution of delays between THE trade (indexes 1-11) and trade-1 (indexes 12-16 in data line).</p>\r\n<p>% each graph cuts that mode in 10 finer slices of time and plots the mean or median absolute error</p>\r\n<p>% the first 5 blobs, which comprise the vast majority of weight, show a strong relationship between time in the blob and abs(price-curve_based)</p>\r\n<p>&nbsp;</p>\r\n<p>% the last two graphs put together all blobs to illustrate the fact that the relationship gets a reset every blob.</p>\r\n<p>&nbsp;</p>\r\n<p>&nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "10724",
      "postDate": "05/01/2012 16:32:28",
      "content": "<p>The attached graph shows the distribution of weight (how much total weight as a function of weight).</p>\r\n<p>obtained by <br>\r\n&gt;&gt; n = hist(train_data(:,4), 1000);<br>\r\n&gt;&gt; n = (1:1000) .* n;<br>\r\n&gt;&gt;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "10725",
      "postDate": "05/01/2012 16:44:23",
      "content": "<p>And these two graphs show, broken down by trade type (2,3,4), the mean and median of abs(target_price - yield_curve_based_price) and &nbsp;abs(target_price - price_last1)</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 10724,
      "author_name": "teaserebotier",
      "author_url": "",
      "post_date": "05/01/2012 16:32:28",
      "content": "<p>The attached graph shows the distribution of weight (how much total weight as a function of weight).</p>\r\n<p>obtained by <br>\r\n&gt;&gt; n = hist(train_data(:,4), 1000);<br>\r\n&gt;&gt; n = (1:1000) .* n;<br>\r\n&gt;&gt;</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 10725,
      "author_name": "teaserebotier",
      "author_url": "",
      "post_date": "05/01/2012 16:44:23",
      "content": "<p>And these two graphs show, broken down by trade type (2,3,4), the mean and median of abs(target_price - yield_curve_based_price) and &nbsp;abs(target_price - price_last1)</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "10710": "",
    "10724": "",
    "10725": ""
  },
  "source": "meta"
}