This research project studies information aggregation problems (defined broadly) in the context of managing marketplaces and online platforms. More specifically, it has three streams.

The first stream studies active learning problems through the lens of sequential experimental design. In an active learning problem, the decision maker (DM) can proactively influence the data flow to make the data aggregation process more efficient. Under a sequential hypothesis testing framework, I strive to develop the general methodologies associated with it.

The second stream aims to study how to apply the active learning methodologies to platform operations, such as preference learning and ranking & selection. A concrete example is to design surveys/questionnaires on an e-commerce platform. I also explore modelling and estimation problems to capture the customer preferences on the platform.

The third stream aims to study dynamic learning and information aggregation problems in the presence of strategic disruptions. Both the theoretical framework and its applications in platform operations are considered, such as fraud clicks and fake orders, among others.