AdequacyModel

AdequacyModel implements a modified Particle Swarm Optimization (PSO) in R to fit probability distributions and provide statistical measures of model adequacy for survival and lifetime data.

Key Features:

• Particle Swarm Optimization (PSO) Technique: Implements a modified Particle Swarm Optimization (PSO) algorithm adapted for complex, flat, or non-differentiable search spaces to enable robust parameter estimation.
• Statistical Measures for Model Adequacy: Provides a comprehensive set of statistical measures to assess the adequacy of fitted probabilistic models.
• Control Over Optimization Process: Offers customizable stop criteria including a minimum number of iterations and variance thresholds among optimal values to control convergence.
• Log-likelihood Maximization: Targets maximization of log-likelihood functions for model fitting where derivative-based methods may fail.
• R package implementation: Distributed as an R package (AdequacyModel v2.0.0) and used in conjunction with Newdistns (v2.1), Wrapped (v2.0), and sglg.

Scientific Applications:

• Probability and Statistics: Fitting complex lifetime distributions and validating probabilistic models.
• Bioinformatics: Applications in survival analysis and probabilistic modeling of biological datasets.
• Physics and Mathematics: Use of optimization techniques for problems in physics and mathematical modeling.
• Computational Statistics: Foundation for related R packages (Newdistns v2.1, Wrapped v2.0) and support for sglg in broader computational statistics workflows.

Methodology:

Leverages a modified PSO algorithm to navigate complex search spaces and maximize log-likelihood functions, addressing flat regions and non-differentiable areas.