Constructing frequentist confidence intervals using the statistical bootstrap

An R Shiny application for causal mediation analysis when mediators are correlated but not causally linked.

Focused on enhancing the reliability of predictive models, this project applies both traditional Bootstrap and Bag of Little Bootstraps (BLB) techniques, optimized for large datasets. We conduct a comparative analysis of BLB's efficiency and accuracy against traditional Bootstrapping technique for several regression and classification models