{
  "id": 286135,
  "title": "How to effectively use the training dataset?",
  "url": "/competitions/g-research-crypto-forecasting/discussion/286135",
  "author_name": "",
  "post_date": "2021-11-08T05:54:24.974749500Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>It seems that there is not time series to process prior to each prediction. Am I missing something?  </p>",
  "messages": [
    {
      "id": "1575042",
      "postDate": "11/08/2021 05:54:24",
      "content": "<p>It seems that there is not time series to process prior to each prediction. Am I missing something?  </p>",
      "rawMarkdown": "It seems that there is not time series to process prior to each prediction. Am I missing something?",
      "votes": null
    },
    {
      "id": "1576011",
      "postDate": "11/08/2021 19:37:37",
      "content": "<p>You have to build up your own data container where you store each new row in. The rows you receive are 1 minute apart. But yes, for the very first rows you don't have much data for making a prediction.</p>",
      "rawMarkdown": "You have to build up your own data container where you store each new row in. The rows you receive are 1 minute apart. But yes, for the very first rows you don't have much data for making a prediction.",
      "votes": null
    },
    {
      "id": "1578010",
      "postDate": "11/10/2021 15:04:55",
      "content": "<p>see the starter notebooks table <a href=\"https://www.kaggle.com/c/g-research-crypto-forecasting/discussion/285726\" target=\"_blank\">here</a></p>",
      "rawMarkdown": "see the starter notebooks table [here](https://www.kaggle.com/c/g-research-crypto-forecasting/discussion/285726)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1576011,
      "author_name": "fritzcremer",
      "author_url": "",
      "post_date": "11/08/2021 19:37:37",
      "content": "<p>You have to build up your own data container where you store each new row in. The rows you receive are 1 minute apart. But yes, for the very first rows you don't have much data for making a prediction.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1578010,
      "author_name": "",
      "author_url": "",
      "post_date": "11/10/2021 15:04:55",
      "content": "<p>see the starter notebooks table <a href=\"https://www.kaggle.com/c/g-research-crypto-forecasting/discussion/285726\" target=\"_blank\">here</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1575042": "It seems that there is not time series to process prior to each prediction. Am I missing something?",
    "1576011": "You have to build up your own data container where you store each new row in. The rows you receive are 1 minute apart. But yes, for the very first rows you don't have much data for making a prediction.",
    "1578010": "see the starter notebooks table [here](https://www.kaggle.com/c/g-research-crypto-forecasting/discussion/285726)"
  },
  "source": "meta"
}