Bar-Crawl-Detecting-Heavy-Drinking

Objective:

To get hands-on experience in implementing time series analysis techniques on real-world datasets

Introduction:

This project aims to analyze the "Bar Crawl: Detecting Heavy Drinking" dataset from the UCI Machine Learning Repository. The initial phase involves understanding the dataset through exploration, preprocessing, and feature analysis. It then explores the use of permutation entropy and complexity methods to differentiate between heavy drinking and sober behavior, aiming to assess their effectiveness in classification. The structured analysis pipeline integrates data preprocessing, feature engineering, model development, and performance evaluation to contribute to advancements in detecting alcohol-related behaviors using computational methods. The study's findings could inform future research on behavioral analysis and intervention strategies.

Data:

https://archive.ics.uci.edu/dataset/515/bar+crawl+detecting+heavy+drinking

Article:

https://ceur-ws.org/Vol-2429/paper6.pdf

Tasks:

1. Understand the data provided for the project
2. Investigate whether permutation entropy and complexity method is reliable in differentiating heavy drinking vs. sober cases