Dual-Model Product Portfolio Optimization Using Econometric and Machine Learning Approaches

Built a hybrid econometric–machine learning portfolio optimization model that uses BLP-based cross-price elasticities and machine learning feature similarity to identify and minimize product cannibalization. This approach resulted in a 20% lift in projected profits and created a more strategically diversified product lineup.