Decision Support in Complex Market Mechanisms

Making decisions in complex market mechanisms often requires one to process large (e.g., exponential) amounts of information efficiently (e.g., in real time). Such decision task is typically well beyond the cognitive capacity of individuals. The goal of this stream of research is to design theory-based computational tools to support economic transactions and decision making in information-intensive markets.


Robust Statistical Inferences with ML-Generated Covariates

Supervised machine learning enables scalable and cost-effective extraction of useful information from (often unstructured) data, which can be subsequently used in econometric estimations. However, the estimations may suffer from bias due to imperfect machine learning predictions. The goal of this stream of research is to understand the nature of such bias, and design novel approaches to mitigate it.


My dissertation research investigates the interplay and relationships between digital content and engagement behaviors in business pages on social media, and the related challenges and opportunities for associated firms. Social media platforms such as Facebook empower individual users to interact with each other on firm-hosted business pages. Users can engage with the content created by the firms as well as by other users in multiple ways (liking, commenting, etc.). Such engagement behavior bears important consequences to business, yet its fundamental characteristics are not well-understood. My dissertation research investigates user engagement behaviors toward user-generated content on Facebook business pages. The dissertation can be accessed here.