A self-generalizing gradient boosting machine that doesn't need hyperparameter optimization - View it on GitHub