MagicTelescope: Simulated Gamma-Ray Shower Images for Particle Classification

Monte Carlo-generated data simulates the registration of high-energy gamma particles in a ground-based atmospheric Cherenkov telescope. The dataset belongs to a benchmark for regression on numerical features and was originally sourced from the UCI Machine Learning Repository. It contains parameters derived from shower image patterns, known as Hillas parameters, used to discriminate gamma-ray signals from cosmic-ray background.

Use Cases

• Benchmarking regression models on numerical features based on the dataset's stated benchmark category.
• Training classifiers to discriminate gamma-ray signals from hadronic background based on described shower image parameters.
• Analyzing simulated particle shower characteristics like ellipse (Hillas) parameters and energy deposition asymmetry.
• Studying feature importance for event classification in Cherenkov telescope imaging data.

Strengths

• Data is generated via the established CORSIKA Monte Carlo simulation program, providing a controlled experimental environment.
• Includes physically meaningful parameters for discrimination, such as Hillas ellipse characteristics and deposition asymmetry.
• Dataset is part of a published tabular data benchmark, suggesting a standardized transformation for comparative analysis.

Limitations

• Row count and column details are unknown, which may limit suitability assessment for specific modeling tasks.
• Column-level documentation is absent; field semantics must be inferred from the general description after download.
• Data is simulated rather than observed, which may not capture all nuances of real-world telescope measurements.

Provenance

Source

UCI Machine Learning Repository, originally from the MAGIC (Major Atmospheric Gamma Imaging Cherenkov) Telescope project.

Collection Method

Generated by the CORSIKA Monte Carlo simulation program to simulate extensive air showers.