{
  "id": 550918,
  "title": "Combination of Linear Regressions as basemodel",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/550918",
  "author_name": "Bharat Swami",
  "post_date": "2024-12-10T08:58:08.071000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi everyone, I am trying to do something new from my side but its getting more and more complicated. It is kind of very small problem and I thought I can solve my own but I can't , so, kindly help me please.</p>\n<p>What I am trying to do?<br>\nBasically, for my base model I am trying to create series of linear regression models which are depend upon the symbol_id. I have trained the models and saved in \"pkl\" file and imported in the notebook.</p>\n<p>One question you may have here which is, what about the different features used for training the respective model. Yes, I have also saved the pandas.index in txt file and also imported in the notebook.</p>\n<p>My Question:<br>\nHow can I combine different models in one model, which when the run on test dataset auto select the model internally which model to use based on the symbol. <br>\nI have tried working with the iterating through the test , which is obviously uses more time.</p>\n<p>Is there any fast way to do this, or should i explore other way, any suggestions will be helpful.</p>",
  "messages": [
    {
      "id": 3068433,
      "postDate": "2024-12-10T08:58:08.070Z",
      "content": "<p>Hi everyone, I am trying to do something new from my side but its getting more and more complicated. It is kind of very small problem and I thought I can solve my own but I can't , so, kindly help me please.</p>\n<p>What I am trying to do?<br>\nBasically, for my base model I am trying to create series of linear regression models which are depend upon the symbol_id. I have trained the models and saved in \"pkl\" file and imported in the notebook.</p>\n<p>One question you may have here which is, what about the different features used for training the respective model. Yes, I have also saved the pandas.index in txt file and also imported in the notebook.</p>\n<p>My Question:<br>\nHow can I combine different models in one model, which when the run on test dataset auto select the model internally which model to use based on the symbol. <br>\nI have tried working with the iterating through the test , which is obviously uses more time.</p>\n<p>Is there any fast way to do this, or should i explore other way, any suggestions will be helpful.</p>",
      "rawMarkdown": "Hi everyone, I am trying to do something new from my side but its getting more and more complicated. It is kind of very small problem and I thought I can solve my own but I can't , so, kindly help me please.\n\nWhat I am trying to do?\nBasically, for my base model I am trying to create series of linear regression models which are depend upon the symbol_id. I have trained the models and saved in \"pkl\" file and imported in the notebook.\n\nOne question you may have here which is, what about the different features used for training the respective model. Yes, I have also saved the pandas.index in txt file and also imported in the notebook.\n\nMy Question:\nHow can I combine different models in one model, which when the run on test dataset auto select the model internally which model to use based on the symbol. \nI have tried working with the iterating through the test , which is obviously uses more time.\n\nIs there any fast way to do this, or should i explore other way, any suggestions will be helpful.\n\n"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3068433": "Hi everyone, I am trying to do something new from my side but its getting more and more complicated. It is kind of very small problem and I thought I can solve my own but I can't , so, kindly help me please.\n\nWhat I am trying to do?\nBasically, for my base model I am trying to create series of linear regression models which are depend upon the symbol_id. I have trained the models and saved in \"pkl\" file and imported in the notebook.\n\nOne question you may have here which is, what about the different features used for training the respective model. Yes, I have also saved the pandas.index in txt file and also imported in the notebook.\n\nMy Question:\nHow can I combine different models in one model, which when the run on test dataset auto select the model internally which model to use based on the symbol. \nI have tried working with the iterating through the test , which is obviously uses more time.\n\nIs there any fast way to do this, or should i explore other way, any suggestions will be helpful.\n\n"
  }
}