{
  "id": 540556,
  "title": "Data Understanding",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/540556",
  "author_name": "",
  "post_date": "2024-10-15T04:35:24.595413300Z",
  "votes": 6,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I have some basic questions about the data structure here. In this context,</p>\n<ol>\n<li>'responders' refers to the financial institutions or investors involved</li>\n<li>'Symbol_id' may refer to the different kinds of financial instruments. </li>\n<li>Lag dataset may refer to the positions that those responders hold at date 0 (with lag 1)<br>\nAm I right?</li>\n</ol>",
  "messages": [
    {
      "id": "3017630",
      "postDate": "10/15/2024 04:35:24",
      "content": "<p>I have some basic questions about the data structure here. In this context,</p>\n<ol>\n<li>'responders' refers to the financial institutions or investors involved</li>\n<li>'Symbol_id' may refer to the different kinds of financial instruments. </li>\n<li>Lag dataset may refer to the positions that those responders hold at date 0 (with lag 1)<br>\nAm I right?</li>\n</ol>",
      "rawMarkdown": "I have some basic questions about the data structure here. In this context,\n1.  'responders' refers to the financial institutions or investors involved\n2. 'Symbol_id' may refer to the different kinds of financial instruments. \n3. Lag dataset may refer to the positions that those responders hold at date 0 (with lag 1)\nAm I right?",
      "votes": null
    },
    {
      "id": "3024222",
      "postDate": "10/21/2024 12:49:01",
      "content": "<p>I have the same basic questions:</p>\n<p>1) What are responders. I try to google it, but did not find a satisfactory answer<br>\n2) Is there some basic literature about machine learning applied in trading to get the context a bit in more detail?</p>",
      "rawMarkdown": "I have the same basic questions:\n\n1) What are responders. I try to google it, but did not find a satisfactory answer\n2) Is there some basic literature about machine learning applied in trading to get the context a bit in more detail?",
      "votes": null
    },
    {
      "id": "3061953",
      "postDate": "12/03/2024 06:39:41",
      "content": "<p>I would suspect something like:</p>\n<table>\n<thead>\n<tr>\n<th>Symbols</th>\n<th>Responders</th>\n<th>Features</th>\n<th>Weights</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>AAPL</td>\n<td>Buy Signal</td>\n<td>Moving Avg., RSI, Volume</td>\n<td>0.4, 0.3, 0.3</td>\n</tr>\n<tr>\n<td>MSFT</td>\n<td>Sell Signal</td>\n<td>Momentum, RSI, News Sentiment</td>\n<td>0.5, 0.2, 0.3</td>\n</tr>\n<tr>\n<td>TSLA</td>\n<td>Hold Signal</td>\n<td>Volatility, News Impact</td>\n<td>0.6, 0.4</td>\n</tr>\n</tbody>\n</table>\n<p>Since train already contains responders data, Lags seems to be responders applicable only to test dataset, at <strong>time 0</strong> for each previous day. The catch is the dates are not consecutive, they are date_ids, so I am not 100% of that join between test and lags.</p>",
      "rawMarkdown": "I would suspect something like:\n\n| Symbols | Responders | Features | Weights |\n| --- | --- | --- | --- | \n| AAPL\t| Buy Signal |\tMoving Avg., RSI, Volume |\t0.4, 0.3, 0.3 |\n| MSFT\t| Sell Signal |\tMomentum, RSI, News Sentiment |\t0.5, 0.2, 0.3 |\n| TSLA\t| Hold Signal |\tVolatility, News Impact |\t0.6, 0.4 |\n\n\nSince train already contains responders data, Lags seems to be responders applicable only to test dataset, at **time 0** for each previous day. The catch is the dates are not consecutive, they are date_ids, so I am not 100% of that join between test and lags.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3024222,
      "author_name": "happykiter",
      "author_url": "",
      "post_date": "10/21/2024 12:49:01",
      "content": "<p>I have the same basic questions:</p>\n<p>1) What are responders. I try to google it, but did not find a satisfactory answer<br>\n2) Is there some basic literature about machine learning applied in trading to get the context a bit in more detail?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3061953,
      "author_name": "icozma",
      "author_url": "",
      "post_date": "12/03/2024 06:39:41",
      "content": "<p>I would suspect something like:</p>\n<table>\n<thead>\n<tr>\n<th>Symbols</th>\n<th>Responders</th>\n<th>Features</th>\n<th>Weights</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>AAPL</td>\n<td>Buy Signal</td>\n<td>Moving Avg., RSI, Volume</td>\n<td>0.4, 0.3, 0.3</td>\n</tr>\n<tr>\n<td>MSFT</td>\n<td>Sell Signal</td>\n<td>Momentum, RSI, News Sentiment</td>\n<td>0.5, 0.2, 0.3</td>\n</tr>\n<tr>\n<td>TSLA</td>\n<td>Hold Signal</td>\n<td>Volatility, News Impact</td>\n<td>0.6, 0.4</td>\n</tr>\n</tbody>\n</table>\n<p>Since train already contains responders data, Lags seems to be responders applicable only to test dataset, at <strong>time 0</strong> for each previous day. The catch is the dates are not consecutive, they are date_ids, so I am not 100% of that join between test and lags.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3017630": "I have some basic questions about the data structure here. In this context,\n1.  'responders' refers to the financial institutions or investors involved\n2. 'Symbol_id' may refer to the different kinds of financial instruments. \n3. Lag dataset may refer to the positions that those responders hold at date 0 (with lag 1)\nAm I right?",
    "3024222": "I have the same basic questions:\n\n1) What are responders. I try to google it, but did not find a satisfactory answer\n2) Is there some basic literature about machine learning applied in trading to get the context a bit in more detail?",
    "3061953": "I would suspect something like:\n\n| Symbols | Responders | Features | Weights |\n| --- | --- | --- | --- | \n| AAPL\t| Buy Signal |\tMoving Avg., RSI, Volume |\t0.4, 0.3, 0.3 |\n| MSFT\t| Sell Signal |\tMomentum, RSI, News Sentiment |\t0.5, 0.2, 0.3 |\n| TSLA\t| Hold Signal |\tVolatility, News Impact |\t0.6, 0.4 |\n\n\nSince train already contains responders data, Lags seems to be responders applicable only to test dataset, at **time 0** for each previous day. The catch is the dates are not consecutive, they are date_ids, so I am not 100% of that join between test and lags."
  },
  "source": "meta"
}