Wine Enthusiast dataset -

Classification models using LDA, Naive Bayes, & text mining

Bayesian Linear Regression models to predict points assigned by wine reviewer

Authors: Will Daniel, Zach Lynch, and John Mark Pittman

FE.py is based on the following git page: https://github.com/gorokhovnik/wine_analysis/

Main files include:

1. points_regression.py - this runs a bayesian regression model on the wine data set using ADVI
2. points_regression_hierarchical.py - this runs a bayesian hierarchical regression model on the wine data set using ADVI
   • In this model we use hyperpriors for each continent on province to create the multi-level model
3. text_mining_with_naive_bayes.ipynb - This creates a dataframe from the description of the wine dataset using text mining
   • Then uses that dataframe to run naive bayes classification models
4. Vino_LDA.ipynb - This produces an LDA model from description & uses the topic proportions to train classification models