{
  "fetched_at": "2026-10-19T21:55:27Z",
  "source_url": "https://www.kaggle.com/competitions/follow-the-trend-time-series-forecasting/data",
  "pages": [
    {
      "name": "data-description",
      "content": "The share data contains aggregated weekly sales of a product in a specific geography.\nEach row in the dataset represents an aggregated transactions quantity.\n\n## Files\n\n*   **train.csv** - the training set\n*   **test.csv** - the test set\n*   **sample_submission.csv** - a sample submission file in the correct format\n\n## Columns\n\n*   `Sales_Date` - Date at which transactions were made\n*   `Product` - Unique identifier of a product\n*   `Country ` -Unique identifier of the Geography\n*   `Sales_Qty ` - Target Column",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "You are invited to participate in this competition where the objective is to develop a sales forecasting model based on sales history, product attributes, and geographic information. The goal is to create a model that can accurately predict sales, even with noisy data input, and effectively follow trends.\n\nThere are no predetermined requirements for which information may be useful in this competition - it is up to you to discover the most effective factors for forecasting. You are free to explore and use any machine learning, deep learning, or statistical models that you believe will be effective in this task.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "Submissions in this competition will be evaluated using the Root Mean Square Error (RMSE) metric, which is the most similar to Weighted Absolute Percentage Error (WAPE).\n\n$$ RMSE = \\sqrt{\\frac{1}{n}\\sum_{t=1}^{n}(A\\_t - F\\_t )^2 }   $$\n\nPlease note that you will be required to make forecasts for 6 future horizons, and you may choose to use either a single model for each horizon or a multi-output model for all horizons.\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "Simple Rules : \n\n - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}