{
  "fetched_at": "2026-10-19T08:33:49Z",
  "source_url": "https://www.kaggle.com/competitions/gpr-data-processing/data",
  "pages": [
    {
      "name": "data-description",
      "content": "## What files do I need?  \n- train_x.npy - input file for train the solution\n- train_y.npy - output file for train the solution\n- test_x.npy - input file for make prediction\n- sample_submission.csv - example of file to make prediction\n\n## What should I expect the data format to be?  \n - I want npy format, but I need to check this\n\n## What am I predicting?  \n- You need to predict 'test_y.csv' file like 'sample_submission.csv' format. It's just a .npy flattened file\n```python\nflat_data = data.reshape(-1)\ndf = pd.DataFrame({'ID Column': np.arange(len(flat_data)), 'Data': flat_data})\ndf.to_csv(csv_path, index=False)\n```\n\n## What acronyms will I encounter?\n-nothing. But if there is great interest in the parameters of simulation and ground penetrating radar, then I will create a separate notebook with a description.\n\n## Files\n\n*   **train_x.npy** - the training input set\n*   **train_y.npy** - the training output set\n*   **test_x.npy** - a sample submission file in the correct format\n*   **sample_submission.csv** - a sample submission file in the correct format",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "# About task\nGPR is one of the geophysical research methods. The essence of the problem comes down to the emission of electromagnetic waves into the near-surface space and obtaining reflections of these waves from electrophysical inhomogeneities.\nThere are different approaches to interpreting the data obtained, but they all have their own shortcomings and inaccuracies. In this competition, it is proposed to test various ML methods for solving a problem to find new solutions for processing georadar data.\n\n# About data\nThe data was generated using the gprMax package. This is synthetic data that has been processed and scaled. The exact values ​​of the simulation here are most likely not interesting, if this is not so, then write.\nThe input signal is a square 128x128x1 - a radio signal, the output signal 128x128x2 - these are the parameters of the environment, dielectric constant and electrical conductivity.\n\n# What do we see in the sample?\n- The output sample is a meter by meter square with a randomly simulated environment with values ​​similar to real ones, which includes metal cylinders distributed randomly. The figure also has a thin layer of air on top.\n- The input signal (x) shows the result of a georadar survey along a profile one meter long (horizontal) and 30 ns deep (vertical).",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The usual MAE is used for all points. The author is looking at whether it is possible to use an npy array for competition",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "No rules. But if you use the data, please reference this competition and gprMax (http://www.gprmax.com) their publications",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "The objective of this competition is to convert ground penetrating radar data into an electrophysical model. Data generated using gprMax in two-dimensional mode, scaled, input data amplified to correct for spherical signal divergence.\nIt is proposed to use ML approaches for the problem, such as: image-to-image, time series and other more classical approaches for wave geophysical problems.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}