Grape Wine Rating Inference

This project aims to leverage machine learning techniques to identify widely acclaimed wines using market-available features, overcoming the limitations of complex, lab-based prediction methods. The core objective is to provide a predictive model that assists three key stakeholders: consumers making purchasing decisions, wine cellar owners selecting inventory to import, and winery owners understanding general public preferences.

collaborator : Kuan-wei Tseng

Project Layout

• src/Xwines.ipynb: main analysis notebook
• reports/Report.pdf: final report artifact
• models/: saved model artifacts
• additional files/: exported OLS and recommendation tables
• data/data_link.txt: source links
• data/README.md: expected local placement for the cleaned Xwines.csv

Local Rerun Notes

To rerun the notebook locally:

1. Download the cleaned Xwines.csv file and place it at data/Xwines.csv.
2. Open src/Xwines.ipynb.
3. The notebook now reads data from data/Xwines.csv, saves model artifacts under models/, and writes table exports under additional files/.

The cleaned CSV stays local and is excluded from Git because of its size.