git clone https://github.com/samuelstevens/buckeyetix-ml.git
cd buckeyetix-ml

pip install -r requirements.txt

git clone https://github.com/samuelstevens/buckeyetix-ml.git
cd buckeyetix-ml

python3 -m venv venv

. venv/bin/activate

pip install -r requirements.txt

Ensure that folder ~/tmp exists (and that you can write/read from it).

cp data/tickets.json ~/tmp/
python src/util/script.py
python src/datasci/linearSVC.py

# one line version (so that it's in bash history)
python src/util/script.py; python src/datasci/linearSVC.py

linearSVC.py seems to be the best balance between not overfitting and still being accurate. gradient.py can reach ~84% with particular parameters (most likely overfitted) and cross_val.py sits around 70% (not as accurate). linearSVC.py uses a LinearSVC with sklearn's StandardScaler() to scale and normalize the data. linearSVC.py achieves about 78% accuracy on both the training data and test data, even when the script.py is run each time (randomizing training vs testing datasets).

• Add row number as parameter
• Optimize process past simply trying every option