{
  "id": 253991,
  "title": "Query regarding pre-processing and feature extraction on the Raw Images",
  "url": "/competitions/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/253991",
  "author_name": "",
  "post_date": "2021-07-19T17:28:54.893122800Z",
  "votes": 2,
  "comment_count": 5,
  "views": 0,
  "content": "<ol>\n<li><p>Can I pre-process the train data and validation data and extract features using MATLAB code?</p></li>\n<li><p>Then I will use those extracted features to train my model and make predictions on test data, can I do that?</p></li>\n<li><p>Or do I have to work with the provided raw data in dicom format and can not do pre-processing outside the notebook of Kaggle platform?</p></li>\n</ol>\n<p>Please assist me with the aforementioned questions.</p>",
  "messages": [
    {
      "id": "1393492",
      "postDate": "07/19/2021 17:28:54",
      "content": "<ol>\n<li><p>Can I pre-process the train data and validation data and extract features using MATLAB code?</p></li>\n<li><p>Then I will use those extracted features to train my model and make predictions on test data, can I do that?</p></li>\n<li><p>Or do I have to work with the provided raw data in dicom format and can not do pre-processing outside the notebook of Kaggle platform?</p></li>\n</ol>\n<p>Please assist me with the aforementioned questions.</p>",
      "rawMarkdown": "1. Can I pre-process the train data and validation data and extract features using MATLAB code?\n\n2. Then I will use those extracted features to train my model and make predictions on test data, can I do that?\n\n3. Or do I have to work with the provided raw data in dicom format and can not do pre-processing outside the notebook of Kaggle platform?\n\nPlease assist me with the aforementioned questions.",
      "votes": null
    },
    {
      "id": "1393595",
      "postDate": "07/19/2021 18:42:06",
      "content": "<p>Same question here!! </p>",
      "rawMarkdown": "Same question here!!",
      "votes": null
    },
    {
      "id": "1393739",
      "postDate": "07/19/2021 21:26:19",
      "content": "<p>As far I know, just the inference notebook has to be I R or python. But is just an assumption</p>",
      "rawMarkdown": "As far I know, just the inference notebook has to be I R or python. But is just an assumption",
      "votes": null
    },
    {
      "id": "1393996",
      "postDate": "07/20/2021 05:10:44",
      "content": "<p>Thank you for your reply.<br>\nI am new to kaggle, could you pleass explain what you mean by inference.<br>\nIn my case, my extracted features from MATLAB will create a csv file and using that csv file I will make the prediction using python.<br>\nDo you think it will work in kaggle platform?</p>",
      "rawMarkdown": "Thank you for your reply.\nI am new to kaggle, could you pleass explain what you mean by inference.\nIn my case, my extracted features from MATLAB will create a csv file and using that csv file I will make the prediction using python.\nDo you think it will work in kaggle platform?",
      "votes": null
    },
    {
      "id": "1394062",
      "postDate": "07/20/2021 06:33:07",
      "content": "<p>That's not going to work. When you press the submit button more pictures are added to the test-folder and you have to make a predcition for all of them. If you just upload a csv that just one part of the game and your going to get a error.</p>\n<p>Siehe here:</p>\n<p><a href=\"https://www.kaggle.com/docs/competitions#notebooks-only-FAQ\" target=\"_blank\">https://www.kaggle.com/docs/competitions#notebooks-only-FAQ</a></p>\n<p>I made a very basic inference notbook, so can get an idea:</p>\n<p><a href=\"https://www.kaggle.com/lucamtb/brain-tumor-very-basice-inference\" target=\"_blank\">https://www.kaggle.com/lucamtb/brain-tumor-very-basice-inference</a> </p>",
      "rawMarkdown": "That's not going to work. When you press the submit button more pictures are added to the test-folder and you have to make a predcition for all of them. If you just upload a csv that just one part of the game and your going to get a error.\n\nSiehe here:\n\nhttps://www.kaggle.com/docs/competitions#notebooks-only-FAQ\n\nI made a very basic inference notbook, so can get an idea:\n\nhttps://www.kaggle.com/lucamtb/brain-tumor-very-basice-inference",
      "votes": null
    },
    {
      "id": "1397931",
      "postDate": "07/23/2021 16:11:34",
      "content": "<p>Basically you have to work with raw data and do all pre-processing inside the notebook. Anyway a solution could be to generate a python package from matlab, upload it on kaggle as a dataset and the install it into your notebook (with pip install).<br>\nI've never done something similar but maybe <a href=\"https://towardsdatascience.com/matlab-function-in-python-739c473c8176\" target=\"_blank\">this</a> can help</p>",
      "rawMarkdown": "Basically you have to work with raw data and do all pre-processing inside the notebook. Anyway a solution could be to generate a python package from matlab, upload it on kaggle as a dataset and the install it into your notebook (with pip install).\nI've never done something similar but maybe [this](https://towardsdatascience.com/matlab-function-in-python-739c473c8176) can help",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1393595,
      "author_name": "ranafago",
      "author_url": "",
      "post_date": "07/19/2021 18:42:06",
      "content": "<p>Same question here!! </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1393739,
      "author_name": "lucamtb",
      "author_url": "",
      "post_date": "07/19/2021 21:26:19",
      "content": "<p>As far I know, just the inference notebook has to be I R or python. But is just an assumption</p>",
      "votes": null,
      "replies": [
        {
          "id": 1393996,
          "author_name": "wfarzana",
          "author_url": "",
          "post_date": "07/20/2021 05:10:44",
          "content": "<p>Thank you for your reply.<br>\nI am new to kaggle, could you pleass explain what you mean by inference.<br>\nIn my case, my extracted features from MATLAB will create a csv file and using that csv file I will make the prediction using python.<br>\nDo you think it will work in kaggle platform?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1394062,
          "author_name": "lucamtb",
          "author_url": "",
          "post_date": "07/20/2021 06:33:07",
          "content": "<p>That's not going to work. When you press the submit button more pictures are added to the test-folder and you have to make a predcition for all of them. If you just upload a csv that just one part of the game and your going to get a error.</p>\n<p>Siehe here:</p>\n<p><a href=\"https://www.kaggle.com/docs/competitions#notebooks-only-FAQ\" target=\"_blank\">https://www.kaggle.com/docs/competitions#notebooks-only-FAQ</a></p>\n<p>I made a very basic inference notbook, so can get an idea:</p>\n<p><a href=\"https://www.kaggle.com/lucamtb/brain-tumor-very-basice-inference\" target=\"_blank\">https://www.kaggle.com/lucamtb/brain-tumor-very-basice-inference</a> </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1397931,
      "author_name": "lazcoder",
      "author_url": "",
      "post_date": "07/23/2021 16:11:34",
      "content": "<p>Basically you have to work with raw data and do all pre-processing inside the notebook. Anyway a solution could be to generate a python package from matlab, upload it on kaggle as a dataset and the install it into your notebook (with pip install).<br>\nI've never done something similar but maybe <a href=\"https://towardsdatascience.com/matlab-function-in-python-739c473c8176\" target=\"_blank\">this</a> can help</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1393492": "1. Can I pre-process the train data and validation data and extract features using MATLAB code?\n\n2. Then I will use those extracted features to train my model and make predictions on test data, can I do that?\n\n3. Or do I have to work with the provided raw data in dicom format and can not do pre-processing outside the notebook of Kaggle platform?\n\nPlease assist me with the aforementioned questions.",
    "1393595": "Same question here!!",
    "1393739": "As far I know, just the inference notebook has to be I R or python. But is just an assumption",
    "1393996": "Thank you for your reply.\nI am new to kaggle, could you pleass explain what you mean by inference.\nIn my case, my extracted features from MATLAB will create a csv file and using that csv file I will make the prediction using python.\nDo you think it will work in kaggle platform?",
    "1394062": "That's not going to work. When you press the submit button more pictures are added to the test-folder and you have to make a predcition for all of them. If you just upload a csv that just one part of the game and your going to get a error.\n\nSiehe here:\n\nhttps://www.kaggle.com/docs/competitions#notebooks-only-FAQ\n\nI made a very basic inference notbook, so can get an idea:\n\nhttps://www.kaggle.com/lucamtb/brain-tumor-very-basice-inference",
    "1397931": "Basically you have to work with raw data and do all pre-processing inside the notebook. Anyway a solution could be to generate a python package from matlab, upload it on kaggle as a dataset and the install it into your notebook (with pip install).\nI've never done something similar but maybe [this](https://towardsdatascience.com/matlab-function-in-python-739c473c8176) can help"
  },
  "source": "meta"
}