Online Demand Learning and Price Optimization with Endogeneity Effect

Project

Online Demand Learning and Price Optimization with Endogeneity Effect

Principal Investigator

David Simchi-Levi, He Wang

Massachusetts Institute of Technology

Oracle Principal Investigator

Sajith Vijayan
Setareh Boroujeni
Su-Ming Wu

Summary

Existing methods used in practice for demand estimation fail to control for price endogeneity, and thus produce inaccurate demand forecast which in turn leads to sub-optimal pricing strategies.

This project involves developing a demand estimation and online price optimization method that produces unbiased price elasticity and improves demand forecast accuracy. The algorithm optimizes the price on the fly, which allows for finding optimal price in the presence of price endogeneity.