{
  "id": 44669,
  "title": "Submission Question",
  "url": "/competitions/passenger-screening-algorithm-challenge/discussion/44669",
  "author_name": "",
  "post_date": "2017-11-30T23:24:07.538071100Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Does the submission need to include predictions for all images? One image, \"42181583618ce4bbfbc0c4c300108bf5\", in particular doesn't make much sense to label as the data was corrupted. There are also two images which have incorrectly swapped labels. Do I really need to include all the training predictions in my submission or can I just submit a file which provides predictions for the stage1 test set? (being images in stage1 that aren't in stage1_labels)</p>",
  "messages": [
    {
      "id": "251286",
      "postDate": "11/30/2017 23:24:07",
      "content": "<p>Does the submission need to include predictions for all images? One image, \"42181583618ce4bbfbc0c4c300108bf5\", in particular doesn't make much sense to label as the data was corrupted. There are also two images which have incorrectly swapped labels. Do I really need to include all the training predictions in my submission or can I just submit a file which provides predictions for the stage1 test set? (being images in stage1 that aren't in stage1_labels)</p>",
      "rawMarkdown": "Does the submission need to include predictions for all images? One image, \"42181583618ce4bbfbc0c4c300108bf5\", in particular doesn't make much sense to label as the data was corrupted. There are also two images which have incorrectly swapped labels. Do I really need to include all the training predictions in my submission or can I just submit a file which provides predictions for the stage1 test set? (being images in stage1 that aren't in stage1_labels)",
      "votes": null
    },
    {
      "id": "251320",
      "postDate": "12/01/2017 01:49:28",
      "content": "<p>You only need the stage1 test set.  See the example submission.</p>",
      "rawMarkdown": "You only need the stage1 test set.  See the example submission.",
      "votes": null
    },
    {
      "id": "253926",
      "postDate": "12/05/2017 21:43:12",
      "content": "<p>could you tell which file is stage1 test set? I thought they are included in the training set, but obviously not.  Thanks.</p>",
      "rawMarkdown": "could you tell which file is stage1 test set? I thought they are included in the training set, but obviously not.  Thanks.",
      "votes": null
    },
    {
      "id": "254489",
      "postDate": "12/07/2017 03:06:30",
      "content": "<p>The stage1 data includes both labeled (training) images and unlabeled (testing) images. Any image in the stage1 dataset which doesn't have a corresponding entry in the stage1_labels file is part of the 'testing' set.</p>",
      "rawMarkdown": "The stage1 data includes both labeled (training) images and unlabeled (testing) images. Any image in the stage1 dataset which doesn't have a corresponding entry in the stage1_labels file is part of the 'testing' set.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 251320,
      "author_name": "hackerpoet",
      "author_url": "",
      "post_date": "12/01/2017 01:49:28",
      "content": "<p>You only need the stage1 test set.  See the example submission.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 253926,
      "author_name": "w9wang2",
      "author_url": "",
      "post_date": "12/05/2017 21:43:12",
      "content": "<p>could you tell which file is stage1 test set? I thought they are included in the training set, but obviously not.  Thanks.</p>",
      "votes": null,
      "replies": [
        {
          "id": 254489,
          "author_name": "",
          "author_url": "",
          "post_date": "12/07/2017 03:06:30",
          "content": "<p>The stage1 data includes both labeled (training) images and unlabeled (testing) images. Any image in the stage1 dataset which doesn't have a corresponding entry in the stage1_labels file is part of the 'testing' set.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "251286": "Does the submission need to include predictions for all images? One image, \"42181583618ce4bbfbc0c4c300108bf5\", in particular doesn't make much sense to label as the data was corrupted. There are also two images which have incorrectly swapped labels. Do I really need to include all the training predictions in my submission or can I just submit a file which provides predictions for the stage1 test set? (being images in stage1 that aren't in stage1_labels)",
    "251320": "You only need the stage1 test set.  See the example submission.",
    "253926": "could you tell which file is stage1 test set? I thought they are included in the training set, but obviously not.  Thanks.",
    "254489": "The stage1 data includes both labeled (training) images and unlabeled (testing) images. Any image in the stage1 dataset which doesn't have a corresponding entry in the stage1_labels file is part of the 'testing' set."
  },
  "source": "meta"
}