{
  "fetched_at": "2026-10-19T10:16:50Z",
  "source_url": "https://www.kaggle.com/competitions/cifar100-image-classification-uaic-fii-atnn-2023/data",
  "pages": [
    {
      "name": "data-description",
      "content": "CIFAR100 Dataset from torchvision.datasets.CIFAR100.\nCheck https://www.kaggle.com/code/onepathsilkworm/notebookdfed7035ac.\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "#### Objective\nStudents are encouraged to explore a variety of techniques, including but not limited to, data preprocessing, data augmentation, test time augmentation (TTA) and hyperparameter finetuning. \n* The model (VGG-16) must not be changed. \n* Only notebooks run and submitted on Kaggle are considered for the 2nd evaluation criteria.  \n#### Evaluation Criteria\nThe homework will be evaluated based on two primary components:\n\n1. **Training experiments (10 points):** \n   - Students must experiment with at least 4 different configurations. (2.5 points each)\n\n2. **Accuracy (10 points):** \n   - Students are graded based on their results. \n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "Submissions are evaluated using the Top-1 and Top-3 accuracy. \nBecause Kaggle does not support two metrics, we will display (top1 + top3) / 2.\n\n## Submission File\nFor each image in the test dataset, include a probability distribution over the 100 classes.\n\n    ID,target\n    0,[0.1, 0.0, 0.9, 0.0, ...]\n    etc.\n\nSee the [sample submission](https://www.kaggle.com/code/onepathsilkworm/notebookdfed7035ac) and [metric calculation](https://www.kaggle.com/code/onepathsilkworm/top-3-top-1-accuracy-score).\n\n",
      "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": "#### Introduction\nWelcome to the CIFAR100 Standard Competition, a core component of the Advanced Topics in Neural Networks class project. In this challenge, students are tasked with developing machine learning models to achieve and surpass the benchmark scores on the CIFAR100 dataset. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Citation",
      "content": "This competition uses the CIFAR 100 dataset: \n* Learning Multiple Layers of Features from Tiny Images, Alex Krizhevsky, 2009. https://www.cs.toronto.edu/~kriz/learning-features-2009-TR.pdf\n* https://www.cs.toronto.edu/~kriz/cifar-100-python.tar.gz\n\n\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}