ATLAS-Higgs-Boson-Machine-Learning-Challenge-2014: Simulated Particle Physics Data

ATLAS-Higgs-Boson-Machine-Learning-Challenge-2014 data originates from the 2014 Kaggle competition and was downloaded from the CERN open data portal. The dataset is used for a binary classification task to distinguish Higgs boson signals from background noise. This version encodes the placeholder value -999 as NaN.

Use Cases

• Train binary classification models to identify Higgs boson decay events based on simulated particle collision features.
• Benchmark machine learning algorithms for physics simulation data using a well-known competition dataset.
• Study feature engineering and preprocessing techniques for handling placeholder values like -999 in scientific datasets.

Strengths

• Data originates from the authoritative CERN open data portal and a major Kaggle competition.
• The dataset has a clear binary classification objective for a significant physics discovery.

Limitations

• Column names and definitions are unknown, requiring users to consult external documentation.
• The total number of rows and features is unknown, making suitability assessment difficult.
• The last update date is unknown, so freshness cannot be verified.

Provenance

Source

CERN open data portal (record 328) and Kaggle Higgs Boson Machine Learning Challenge 2014.

Collection Method

Simulated particle collision data from the ATLAS experiment.