Using a trained Python model to determine if input variables yield a positive or negative result

Question

I am currently using a scikit-learn Stochastic Gradient Booster (SGB) model. The code runs well in terms of tranining and testing the data while returning an Root Mean Square Er