Industrial engineering professor Shipra Agrawal wins NSF CAREER Award
She will develop new learning frameworks and algorithms to account for consumer behavior in product pricing and recommendation
Credit: Timothy Lee Photographers
New York, NY–March 04, 2019–Shipra Agrawal, assistant professor of industrial engineering and operations research at Columbia Engineering, has won a National Science Foundation CAREER Award for her research, “Consumer Behavior-Aware Learning for Revenue Management.” One of the NSF’s most prestigious honors for faculty at or near the beginning of their careers, the five-year, $500,000 award will fund research Agrawal aims to conduct in collaboration with industry partners.
Agrawal, who joined Columbia Engineering in 2015, works at the intersection of operations research and data science. She combines optimization and learning techniques to design globally optimal decision-making frameworks for large, complex, and uncertain environments.
“I am deeply grateful for this recognition of my research by the NSF,” said Agrawal. “I hope my project will have far-reaching implications for online marketplaces and benefit both consumers and sellers through a more efficient matching of supply and demand.”
Modern consumer habits have been dramatically reshaped by the ubiquity of online customer review platforms, price aggregators, discussion forums, and connected communities. Today, a shopper increasingly forms her opinions based on social influences, imitates other users, learns and anticipates prices, and times her purchases. Market mechanisms that fail to account for such behavior can result in significant losses in revenue and inventory mismatches for businesses, as well as negatively affect the consumer’s welfare.
In her research, Agrawal will investigate new learning techniques for dynamic pricing, product recommendations, and allocation under three broad types of commonly occurring consumer behavior: social influence and imitation, forward-looking and planning behavior, and consumer learning. In each category, she will develop new models and decision-making frameworks that can capture natural and uncertain consumer behavior, as well as learn from it. She will also develop new methods for exploration-exploitation, dynamic learning, and online optimization that will build upon and advance tools from multi-armed bandits, online convex optimization, dynamic programming, and reinforcement learning.
Ultimately, she aims to not only provide rigorous theoretical analysis, but also to test her theories and models by working with industry partners to evaluate performance in actual markets.