{
  "fetched_at": "2026-10-19T07:19:32Z",
  "source_url": "https://www.kaggle.com/competitions/col-774-autumn-2020/data",
  "pages": [
    {
      "name": "data-description",
      "content": "<p>  Data Description </p>\n<h2>See the assignment pdf</h2>\n<ul>\n</ul>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "In this assignment, we will deal with the problem of identifying a person’s emotional state through an imageof his/her facial features.  The dataset comes from a set of 7 different labels (emotional states).  Each image isof size 48×48 pixels and containing only 1 channel(greyscale).Note that the dataset is imbalanced hence we will use Macro-F1 as the evaluation metric instead of accuracy.\n\nTry out algorithms of your choice to maximize the accuracy on the test data.  Improve your model as muchas you can.  You are strongly encouraged to try out variations of the algorithms described in the class as well as try new algorithms/models by looking up online resources.\n\nYou will submit the labels for the test dataset on Kaggle and your performance (on classification accuracy) will be ranked on the leader-board. There are 2 leader-boards;  a Public leader-board on which you canmonitor your accuracy on 30% of the test dataset;  a Private leader-board (which you cannot see) which will be used to determine the final rankings on the other 70% of the test data.\n\nNote:  The  score  obtained  in  this  part  will  solely  depend  on  the  quality  of  your  predictions  on  the  test labels.  You must submit the code for your best performing model (as well as include all the details in your report) so that we can replicate the results if required.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "We will use Macro-F1 as the evaluation metric.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "This is a page where you can include rules that participants must accept before joining. You may wish to include rules like:\n\n - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}