A/B/n Experiment Project using Response Surface Methodology
This is a data science project that was done as part of the MSDS 629 - Experiments in Data Science course.
This project aims to optimize a content streaming platform's homepage by minimizing the average browsing time a user takes before watching actual content.
The project uses Response Surface Methodology (RSM) to perform a series of experiments with four design factors: Tile Size, Match Score, Preview Length, and Preview Type.
The project reports the optimal values for these factors that minimize the average browsing time.
You may find the various experiments that were performed in the notebooks that are present in this repo.
raw_data contains the different experimental conditions that were sampled and the corresponding response estimates recorded for each experiment.