{
  "id": 93626,
  "title": "more than 4194 samples?",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/93626",
  "author_name": "",
  "post_date": "2019-05-28T19:13:50.505502800Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Most of kernels generated 4194 samples from the original training set. Each sample is followed by another sample. Why not generate some overlapping samples, which will have more samples to train? Does anyone try? I guess it will increase performance since it has more training samples.</p>",
  "messages": [
    {
      "id": "538571",
      "postDate": "05/28/2019 19:13:50",
      "content": "<p>Most of kernels generated 4194 samples from the original training set. Each sample is followed by another sample. Why not generate some overlapping samples, which will have more samples to train? Does anyone try? I guess it will increase performance since it has more training samples.</p>",
      "rawMarkdown": "Most of kernels generated 4194 samples from the original training set. Each sample is followed by another sample. Why not generate some overlapping samples, which will have more samples to train? Does anyone try? I guess it will increase performance since it has more training samples.",
      "votes": null
    },
    {
      "id": "538605",
      "postDate": "05/28/2019 20:45:14",
      "content": "<p>There are several kernels where participants have tried this:</p>\n\n<p><a href=\"https://www.kaggle.com/zikazika/useful-new-features-and-a-optimised-model\">https://www.kaggle.com/zikazika/useful-new-features-and-a-optimised-model</a></p>\n\n<p><a href=\"https://www.kaggle.com/ricarddelgado/lanl-sampling-schemes\">https://www.kaggle.com/ricarddelgado/lanl-sampling-schemes</a></p>",
      "rawMarkdown": "There are several kernels where participants have tried this:\n\nhttps://www.kaggle.com/zikazika/useful-new-features-and-a-optimised-model\n\nhttps://www.kaggle.com/ricarddelgado/lanl-sampling-schemes",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 538605,
      "author_name": "superluminal098",
      "author_url": "",
      "post_date": "05/28/2019 20:45:14",
      "content": "<p>There are several kernels where participants have tried this:</p>\n\n<p><a href=\"https://www.kaggle.com/zikazika/useful-new-features-and-a-optimised-model\">https://www.kaggle.com/zikazika/useful-new-features-and-a-optimised-model</a></p>\n\n<p><a href=\"https://www.kaggle.com/ricarddelgado/lanl-sampling-schemes\">https://www.kaggle.com/ricarddelgado/lanl-sampling-schemes</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "538571": "Most of kernels generated 4194 samples from the original training set. Each sample is followed by another sample. Why not generate some overlapping samples, which will have more samples to train? Does anyone try? I guess it will increase performance since it has more training samples.",
    "538605": "There are several kernels where participants have tried this:\n\nhttps://www.kaggle.com/zikazika/useful-new-features-and-a-optimised-model\n\nhttps://www.kaggle.com/ricarddelgado/lanl-sampling-schemes"
  },
  "source": "meta"
}