{
  "fetched_at": "2026-10-20T02:18:53Z",
  "source_url": "https://www.kaggle.com/competitions/2024-fitting-a-3d-3d-function-with-deep-learning/data",
  "pages": [
    {
      "name": "data-description",
      "content": "Try to fit the R^3 -> R^3 five dimensional function with a neural network.\nYou are free to choose the architecture and training method you want.\n\nThe samples you have access to have NOT been noised.\n\n\n## Files\n\n*   **3D_3D.csv** - the training set\n*   **3D_3D-test_set.csv** - the test set\n*   **3D_3D-submission_example.csv** - a sample submission file in the correct format",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "f maps elements from 3D real number space to 3D real number space.\nTraining examples are Identically Normally Distributed in R^3, as well as the testing.\nThe test is a sum of absolute errors (L1 norm).\n\n## Marking scheme:\n- <10 000 000 -> 50%\n- <5 000 000 -> 60%\n- <1 000 000 -> 70%\n- <500 000 -> 80%\n- <100 000 -> 90%\n- <50 000 -> 100%",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "Your anonymized information may be utilized for conducting basic statistical analysis and for the purpose of training an ensemble learning during a TD. It is also possible that the information may be included in a scholarly publication.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}