https://github.com/sshleifer/nbadraft
Predicting NBA performance using college box-score stats and combine measurements.
https://github.com/sshleifer/nbadraft
Last synced: about 1 year ago
JSON representation
Predicting NBA performance using college box-score stats and combine measurements.
- Host: GitHub
- URL: https://github.com/sshleifer/nbadraft
- Owner: sshleifer
- Created: 2014-06-28T23:43:08.000Z (almost 12 years ago)
- Default Branch: master
- Last Pushed: 2014-08-06T04:00:32.000Z (almost 12 years ago)
- Last Synced: 2025-04-14T06:13:04.586Z (about 1 year ago)
- Language: Python
- Size: 4.02 MB
- Stars: 7
- Watchers: 2
- Forks: 6
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
##nbaDraft
Predicting NBA performance with college box-score stats, combine measurements and python.
###datasets
The data used to make predictions is in colPro.csv, comprised from draft_mix.csv, colData.csv, bestRapm.csv and measurements.
###scrapers
A handful of web requests that make the datasets you see in the datasets folder.
They take ~1hr to run, so I'll update them when they need to be updated.
###assemble_dataset.py
Merges many of the datasets into colPro.csv, and then splits that into train.csv and test.csv.
###model.py
My attempts at using NCAA boxscore stats and measurements to predict Regularized Adjusted Plus Minus for former US college players. I try different techniques for feature engineering, missing data, and ensemble modeling.