Statistical Model Selection Guide

Recommends the most appropriate statistical or machine learning modeling approaches for a described prediction problem — matching regression, survival, tree-based, and probabilistic methods to the target type, dataset characteristics, interpretability constraints, and regulatory context. Delivers a ranked shortlist of 3-5 candidate models with rationale, assumptions, validation strategy, and a fallback if the primary approach fails. Use when deciding which model to build for a prediction or risk problem, comparing modeling approaches, or justifying a method choice to stakeholders or regulators.