-
Proposed a novel estimator for calculating the variance in asymptotic normal distribution of the Monte Carlo Standard Error (MCSE) using auto-regressive processes (1) for univariate and multivariate case
-
Performed comparative analysis with the batch means estimator by running simulations in R on the data generated using the Metropolis Hastings algorithm with a Bayesian Logistic Regression model fitted to the titanic dataset, and concluded better significantly performance with low batch size and significantly correlated data
This repo contains code for the simulation as well as other utility code for production of graphs and relevant csvs.