← Back to work
Football · live repo
Live repoOpen data (StatsBomb)Contextual Football Metrics
CxG / CxA / CxT training pipeline
View the repo →What this is
A single Prefect flow under one MLflow parent run, with a model ladder from baseline through GLM to boosting and neural, and a from-scratch graph-attention encoder in plain PyTorch.
Metrics
Contextual GLM
0.8131 AUC · Brier 0.06669
vs logistic 0.7982
Validation
5-fold CV 0.8083
Testing
560 tests
Stack
PrefectMLflowPyTorch560 tests
What I would not claim
- catboost
- captum
- plotly
- kaleido
- pydantic
- imbalanced-learn
- DVC-tracked pipeline
- CI