This is the repo in accordance with these three Medium articles:
- Applied Bayesian Inference with PyMC3 pt.1
- Applied Bayesian Inference with PyMC3 pt.2
- Applied Bayesian Inference with PyMC3 and Bambi pt.3
In part 1, I introduce modeling the conditional world via Python/PyMC3 from a contrived coin flip example. In part 2, I dive deeper into Bayesian Analysis with a dataset from Kaggle. StockX's 2019 Sneaker dataset can be found here: https://www.kaggle.com/hudsonstuck/stockx-data-contest. In part part 3, I take the skills learned so far to build ML models based on Bayesian estimation to predict streams for Spotify's Top 200 songs. The dataset can be found here: https://www.kaggle.com/sashankpillai/spotify-top-200-charts-20202021
Part 1 Contents:
- Thinking Bayes
- Probabalistic Programming
Part 2 Contents:
- Introduction
- Exploratory Data Analysis & Data Cleaning
- Modeling & Analysis
- Group Comparison
- Conclusion
Part 3 Contents:
- Introduction
- Exploratory Data Analysis & Data Cleaning
- The Simple Regression
- Robust Regression and Out-of-Sample Prediction
- Multiple, Hierarchical, and Generalized Linear Models
- Model Compare
- Conclusion