Beta Kernel Density Estimator Bandwidth Selector with 35,000x Speedup

18.4 MB of supporting materials for a fast, closed-form bandwidth selector for the beta kernel density estimator. The dataset includes a Python package and documentation authored by Johan Hallberg Szabadváry, last updated on 2026-05-08. The method reduces bandwidth selection complexity from iterative optimization to O(1) and is validated on socioeconomic data.

Use Cases

• Density estimation for socioeconomic indicators based on the described validation on real-world socioeconomic data
• Benchmarking bandwidth selection methods based on the Monte Carlo simulation results mentioned in the description
• Implementing boundary-bias-free density estimation for unit interval data based on the beta kernel's theoretical superiority described
• Replacing Gaussian kernel estimators in standard tools based on the provided drop-in replacement Python package

Strengths

• Includes a comprehensive, open-source Python package for immediate adoption
• The derived Beta Reference Rule delivers a speedup of over 35,000 times compared to numerical optimization
• Real-world validation shows the method avoids common artifacts like 'vanishing boundary' and 'shoulder' issues

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• The dataset's primary content appears to be a method description and software package rather than a traditional data table

Provenance

Source

figshare

Collection Method

Derived from theoretical work and validated with Monte Carlo simulations and real-world socioeconomic data.

Freshness

Last updated 2026-05-08 16:52:01; freshness should be verified