In this part of the teaching material, I would like to present you with a graduate course module on Adaptive Experiments Design we deployed at Carnegie Mellon University. The module is deployed in the following Interactive Data Science Course (HCI Graduate: 05-839), I want to highlight how I am connecting my ongoing research on adaptive experimentation with possible graduate level course design:

This RPub page is the code base for simulation results of the short paper track submission of WAPTS: A Allocation Probability adjusted Thompson Sampling Algorithm for Sparse and Adaptive Contextual Bandits submission, the code has three parts: 1. the implementation of Contextual Thompson Sampling and WAPTS using a Laplace theorem based Gradient descent estimation of the posterior estimation process. 2. plot of the reward setting under different effect sizes, for both policies, for a sample size of 300. 3. plot of the regret setting under different effect sizes, for both policies, for a sample size of 300. Feel free to let us know if there is any confusion!

re-creation on binfd power analysis

This is the code base for all the simulation results I had for the oral examination at the department of statistical science in University of Toronto

Homework XD for the bayesian simulation