Bayesian A/B Testing and building an interactive widgets using ipywidgets

Tools used: python 3.9, numpy, pandas, math, scipy, random, matplotlib, seaborn and ipywidgets

Table of Contents:

• Bayesian A/B Testing
• Monte Carlo Simulation
• Interactive web application

Step by step code implementation:

1. Two variants of emails are sent to users. One without and one with image to randomly picked users
2. Beta distribution (continuous) has been utilise to identifying the probabilities of seeing a particular number of success (clicked) and failures (do not click)
3. The bayesian A/B testing simulated using a Monte Carlo Simulation to check in how many worlds the B variant is better than A variant.
4. Calculated how much better is each B variant than each variant A.
5. Createded an interactive app using ipywidgets to make the bayesian A/B testing interactive.