rFSA

rFSA provides enhanced variable selection for statistical regression models in R by systematically searching model spaces to identify models that optimize user-specified model-quality criteria.

Key Features:

• Algorithmic Efficiency: Employs an algorithm capable of handling large datasets and exploring complex subsets including higher-order interaction terms.
• Feasible Solutions Generation: Generates a set of candidate models through numerous iterations for subsequent evaluation.
• Model Flexibility: Supports linear and generalized linear models for varied regression tasks.
• Criterion Functions: Accommodates criterion functions such as Allen's PRESS and AIC and accepts user-specified model-quality criteria.

Scientific Applications:

• Bioinformatics and healthcare research: Assists in formulating and refining regression models for large-scale datasets with complex interactions.

Methodology:

Uses an iterative algorithmic search to explore potential solutions (including higher-order interaction terms) and evaluates candidate models against criterion functions such as Allen's PRESS and AIC.