{
  "id": 154817,
  "title": "Input features in tfrec files",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/154817",
  "author_name": "",
  "post_date": "2020-05-30T00:52:43.798807300Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I decoded one tfrec file it is having only three features \"image\" ,\"image_data\", \"target\",\nAs the tfrec file contains only the image data in it . But the model was based on more params than only the image data. \"train.csv\" contains More features \nsex - the sex of the patient (when unknown, will be blank)\nage_approx - approximate patient age at time of imaging\nanatom_site_general_challenge - location of imaged site\ndiagnosis - detailed diagnosis information (train only)\nbenign_malignant  - indicator of malignancy of imaged lesion\nthat can be trained on.How can we train the model using multiple inputs ? \nSurely it can improve the model perfomance on test dataset</p>",
  "messages": [
    {
      "id": "867056",
      "postDate": "05/30/2020 00:52:43",
      "content": "<p>I decoded one tfrec file it is having only three features \"image\" ,\"image_data\", \"target\",\nAs the tfrec file contains only the image data in it . But the model was based on more params than only the image data. \"train.csv\" contains More features \nsex - the sex of the patient (when unknown, will be blank)\nage_approx - approximate patient age at time of imaging\nanatom_site_general_challenge - location of imaged site\ndiagnosis - detailed diagnosis information (train only)\nbenign_malignant  - indicator of malignancy of imaged lesion\nthat can be trained on.How can we train the model using multiple inputs ? \nSurely it can improve the model perfomance on test dataset</p>",
      "rawMarkdown": "I decoded one tfrec file it is having only three features \"image\" ,\"image_data\", \"target\",\nAs the tfrec file contains only the image data in it . But the model was based on more params than only the image data. \"train.csv\" contains More features \nsex - the sex of the patient (when unknown, will be blank)\nage_approx - approximate patient age at time of imaging\nanatom_site_general_challenge - location of imaged site\ndiagnosis - detailed diagnosis information (train only)\nbenign_malignant  - indicator of malignancy of imaged lesion\nthat can be trained on.How can we train the model using multiple inputs ? \nSurely it can improve the model perfomance on test dataset",
      "votes": null
    },
    {
      "id": "867188",
      "postDate": "05/30/2020 04:36:56",
      "content": "<p>use it as an additional feature which training your images</p>",
      "rawMarkdown": "use it as an additional feature which training your images",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 867188,
      "author_name": "pratikasarkar",
      "author_url": "",
      "post_date": "05/30/2020 04:36:56",
      "content": "<p>use it as an additional feature which training your images</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "867056": "I decoded one tfrec file it is having only three features \"image\" ,\"image_data\", \"target\",\nAs the tfrec file contains only the image data in it . But the model was based on more params than only the image data. \"train.csv\" contains More features \nsex - the sex of the patient (when unknown, will be blank)\nage_approx - approximate patient age at time of imaging\nanatom_site_general_challenge - location of imaged site\ndiagnosis - detailed diagnosis information (train only)\nbenign_malignant  - indicator of malignancy of imaged lesion\nthat can be trained on.How can we train the model using multiple inputs ? \nSurely it can improve the model perfomance on test dataset",
    "867188": "use it as an additional feature which training your images"
  },
  "source": "meta"
}