xhh6fb/ds4320-project-1: ds4320-project-1
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This release contains the completed materials for DS 4320 Project 1: Predicting NFL Pregame Home-Team Win Probability. The project builds a relational secondary dataset from NFL schedule and team data accessed through nflreadpy, then transforms those data into four linked tables: teams, games, team_games, and matchups. The final dataset is exported as parquet files and documented through a full data table, data dictionary, and uncertainty discussion. This release also includes a reproducible analysis pipeline in Python, Markdown, and SQL. The pipeline loads the final parquet tables into DuckDB, runs SQL queries for exploratory analysis, and trains two classification models — logistic regression and random forest — to predict whether the home team will win using only pregame information. The final selected model is logistic regression, chosen using validation ROC AUC, and the repository includes visualizations, a press release, background readings, and all supporting documentation needed to reproduce the project workflow.
Fields of Study & Subjects
Sports Analytics and Performance
Data Mining and Machine Learning Applications
Forecasting Techniques and Applications
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 1 April 2026
- Language
- EN