{
  "fetched_at": "2026-10-19T17:13:20Z",
  "source_url": "https://www.kaggle.com/competitions/feelings/data",
  "pages": [
    {
      "name": "data-description",
      "content": "## Data\nIn this section you will find data needed to train your NN and make the best prediction.\n\n## Files\n**plantilla_howyoufeel_images.ipynb** jupyter notebook to guide in your first Images Competition .😎\n**sample_submission.csv:** your submission should be something like this, same columns, same shape, same id, different emotions 💪\n**train:** this directory contains two other directories happy😃 and sadness☹️ which contains the images.\n**test:** this directory contains the pictures which you have to predict the emotions of.\n\n**train_set.csv:** on this file you'll find the labels related to the path where the pictures are.\n**test_set.csv:** this file should help you to make the predictions and prepare your submission file.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "###  Facial Expression Recognition\n\nGuess whether these people are feeling good or not regarding their faces. 👍 👎\n\nIt is believed that a human has at least 8 basics emotions, where sadness and happiness are the ones the people usually talk more about. For the human being is kind of an easy job to find out how a person is feeling by the expression on their face. But this time you should train a Neural Network, using the layers and params you may find necessaries in order to get the highest accuracy, differentiating between sad and happy people.\n\nYou're right, people can feel more than one emotion at a time but, let's forget about it for a sec and try to simplify the problem.\n\nThe pictures are already in black and white, have a 48x48 size and have the faces cropped, what should makes everything a little easier. However, feel free to make any changes to the pictures, but remember to make the same kind of treatment to the train set of images and the test set of images, just before you make the predictions.\n\nIn case you don't already know, this is a problem of classification, with two classes.\n\nAre you happy? Are you sad?",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "Your submissions will be scored as the percentage of correct predictions. \n**Categorization accuracy**🏅  Submit results \nThe test set is divided into a public part (with which the public leaderboard is calculated, accessible during the competition) and another private part, with which the final positions are calculated, after the end of the competition. \nPredictions will be sent in the format indicated in the sample_submission.csv file in the data section.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "- Keep calm and code!!\n- Don't cheat!\n- Apply yourself!\n- Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}