{
  "id": 443118,
  "title": "Struggling with LSTM",
  "url": "/competitions/rsna-2023-abdominal-trauma-detection/discussion/443118",
  "author_name": "Alex Luna",
  "post_date": "2023-09-25T15:44:12.798000",
  "votes": 2,
  "comment_count": 1,
  "views": 0,
  "content": "<p>There is someone who is implementing LSTM? I want to use LSTM arquitecture for making predictions but i am lost on how to determine the correct preprocess of the data, i want to use a CNN-LSTM approach, i want to make the features that the CNN gets and input that to the LSTM but I want something to be sure that i am inputing the deep feature vectors in the correct format and using all the benefits of the LSTM arquitecture. Can someone guide me? I have already trained a model but it overfits🥲</p>",
  "messages": [
    {
      "id": 2455606,
      "postDate": "2023-09-25T15:44:12.800Z",
      "content": "<p>There is someone who is implementing LSTM? I want to use LSTM arquitecture for making predictions but i am lost on how to determine the correct preprocess of the data, i want to use a CNN-LSTM approach, i want to make the features that the CNN gets and input that to the LSTM but I want something to be sure that i am inputing the deep feature vectors in the correct format and using all the benefits of the LSTM arquitecture. Can someone guide me? I have already trained a model but it overfits🥲</p>",
      "rawMarkdown": "There is someone who is implementing LSTM? I want to use LSTM arquitecture for making predictions but i am lost on how to determine the correct preprocess of the data, i want to use a CNN-LSTM approach, i want to make the features that the CNN gets and input that to the LSTM but I want something to be sure that i am inputing the deep feature vectors in the correct format and using all the benefits of the LSTM arquitecture. Can someone guide me? I have already trained a model but it overfits🥲",
      "votes": 2
    },
    {
      "id": 2457912,
      "postDate": "2023-09-27T08:29:09.403Z",
      "rawMarkdown": "",
      "votes": -7,
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 2457912,
      "author_name": "",
      "author_url": "",
      "post_date": "2023-09-27T08:29:09.403000",
      "content": "",
      "votes": -7,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2455606": "There is someone who is implementing LSTM? I want to use LSTM arquitecture for making predictions but i am lost on how to determine the correct preprocess of the data, i want to use a CNN-LSTM approach, i want to make the features that the CNN gets and input that to the LSTM but I want something to be sure that i am inputing the deep feature vectors in the correct format and using all the benefits of the LSTM arquitecture. Can someone guide me? I have already trained a model but it overfits🥲",
    "2457912": ""
  }
}