HEPMASS Particle Classification Dataset

HEPMASS is a dataset for classifying particle collisions as signal or background. It contains simulated data from high-energy physics experiments, specifically designed for machine learning applications. The dataset was created by researchers and is hosted on the UCI Machine Learning Repository.

Use Cases

• Classify collision events as signal or background using kinematic features like momentum and energy.
• Benchmark binary classification algorithms on a high-dimensional dataset with 28 features.
• Analyze feature importance for particle identification, using attributes such as particle mass and transverse momentum.
• Study class imbalance handling techniques, as the background class significantly outnumbers the signal class.

Strengths

• Dataset contains over 10 million simulated particle collision events.
• Provides 28 numerical features per event for detailed analysis.

Limitations

• Data is simulated, not from real-world detector observations, which may limit real-world applicability.
• Class distribution is imbalanced, with the background class being more prevalent.

Provenance

Source

UCI Machine Learning Repository

Collection Method

Simulated data from high-energy physics Monte Carlo generators.