{
  "id": 295007,
  "title": "Dimension Reduction and Regression",
  "url": "/competitions/cs5785-fall-2021-final/discussion/295007",
  "author_name": "",
  "post_date": "2021-12-13T23:57:08.544980500Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>If anyone is still working, my team got good results with the cross decomposition methods for dimension reduction and regression. Some of the sklearn models are Partial Least Squares Regression and Partial Least Squares Regression SVD. We got significantly better results using this than performing PCA and then a regression model on the reduced data. </p>",
  "messages": [
    {
      "id": "1617197",
      "postDate": "12/13/2021 23:57:08",
      "content": "<p>If anyone is still working, my team got good results with the cross decomposition methods for dimension reduction and regression. Some of the sklearn models are Partial Least Squares Regression and Partial Least Squares Regression SVD. We got significantly better results using this than performing PCA and then a regression model on the reduced data. </p>",
      "rawMarkdown": "If anyone is still working, my team got good results with the cross decomposition methods for dimension reduction and regression. Some of the sklearn models are Partial Least Squares Regression and Partial Least Squares Regression SVD. We got significantly better results using this than performing PCA and then a regression model on the reduced data.",
      "votes": null
    },
    {
      "id": "1617284",
      "postDate": "12/14/2021 02:26:51",
      "content": "<p>Given there was still time left, decided to try this out for feature reduction and it did help our score by a couple of percentage points. </p>",
      "rawMarkdown": "Given there was still time left, decided to try this out for feature reduction and it did help our score by a couple of percentage points.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1617284,
      "author_name": "petergribizis",
      "author_url": "",
      "post_date": "12/14/2021 02:26:51",
      "content": "<p>Given there was still time left, decided to try this out for feature reduction and it did help our score by a couple of percentage points. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1617197": "If anyone is still working, my team got good results with the cross decomposition methods for dimension reduction and regression. Some of the sklearn models are Partial Least Squares Regression and Partial Least Squares Regression SVD. We got significantly better results using this than performing PCA and then a regression model on the reduced data.",
    "1617284": "Given there was still time left, decided to try this out for feature reduction and it did help our score by a couple of percentage points."
  },
  "source": "meta"
}