About

This comprehensive Titanic survival analysis canvas executes a complete exploratory data analysis and machine learning workflow, combining multiple statistical and ML techniques to identify survival drivers. The workflow flows from data loading and preprocessing through exploratory analysis (distributions, correlations, missing values) into feature engineering, then branches into six parallel analytic streams: tree-based importance (Random Forest & Gradient Boosting), permutation importance, logistic regression (odds ratios), statistical tests (chi-square & t-tests), and advanced pattern discovery (anomalies, interactions, clustering). All findings converge in a comprehensive ranking block that aggregates importance scores across methods using normalized composite scoring to definitively rank survival predictors.