{
  "fetched_at": "2026-10-19T08:45:36Z",
  "source_url": "https://www.kaggle.com/competitions/fall-2023-cisc6000-deep-learning-hw1/data",
  "pages": [
    {
      "name": "data-description",
      "content": "mnist.test.npy: Testing images in numpy array.\n\nmnist.train.npy: Training images in numpy array.\n\nmnist.trainlabel.npy: The labels of training images in numpy array.\n\nsample_submission.csv: Your submission format should be like this, the first column is the test image ID, and the second column is the category of your prediction.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "**Image Classification, multiclass.**\n\n- In Kaggle:\n\n  Submit your CSV file of Q3(c) only\n\n  Download your latest version of all code in notebook format including Q2 and Q3\n\n  Download all your CSV files including Q2 and Q3 in the output folder\n\n- In Blackboard:\n\n  Submit your answer to Q1 in a pdf file.\n\n- Q2:\n\n  (1) Submit a CSV file of test data prediction\n\n  (2) Submit your all code in notebook format\n\n- Q3:\n\n  (1) Submit two CSV files of test data prediction\n\n  (2) Submit your all code in notebook format\n\n  (3) Submit Table 1 in a pdf file",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": " - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "- Build Your Own Neural Network\n\n- Tensorflow/Keras to the Rescue",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}