Supervised machine learning techniques, including ensemble tree and other non parametric techniques for prediction of binary and continuous outcomes, e.g. xgboost, LightGBM, CatBoost, deep learning
Feature engineering, including feature ideation, and creation via transformations and aggregations of raw data
Model development process, including train/test, k-fold CV, appropriate performance metrics
Unsupervised methods for clustering and segmentation of consumers, including k-means, k-mode, DBSCAN
Experience translating business problems into supervised and unsupervised machine learning problems to find solutions
Optimization and experimentation of outcomes with multiple controllable parameters, including use of simulations, approximations and assumptions as appropriate
#J-18808-Ljbffr