Description

We built, compared, and optimized various Machine Learning models such as logistic regression and XGBoost to pinpoint relevant factors conducive to predicting couriers’ decisions and behavior, and narrowed down the error of delivery time prediction to around 1.5 minutes.