Abstract: New online platforms are profoundly altering our social and economic interactions by empowering new marketplaces and new collective behavior at increasingly larger scales. Many of these platforms have developed organically, and are not optimally designed for their purpose; furthermore, they are reaching scales where theoretical modeling and operational optimization can have a significant impact.
In this talk, I will highlight two such disruptive technologies - (1) collaborative research, and (2) ride-sharing. In the first part, I will try to uncover the strategic motivations of agents involved in large multi-task research projects; our main finding here is that in a wide variety of settings, having `locally-fair' rewards, proportional to task-difficulty, also indirectly promotes information-sharing. In ride-sharing platforms, I will explore the value of fast timescale dynamic pricing - where prices change every minute, reacting to the instantaneous system state. I will argue that this may not be better than the optimal quasi-static price, but is much more robust to fluctuations in the system parameters. In both settings, the insights I present are reliant on new models we develop for reasoning about the operations of the platform, combining stochastic models for the platform dynamics with with game theoretic models of agent behavior.
Joint work with Ashish Goel, Ramesh Johari, Anilesh Krishnaswamy, Carlos Riquelme and the Data Science team at Lyft.
Bio: Sid Banerjee is a postdoctoral researcher in the Department of Management Science and Engineering at Stanford. He received a PhD in Electrical and Computer Engineering from the University of Texas at Austin in 2013. He is interested in stochastic modeling and the design of algorithms and incentives for large-scale settings, with applications in matching markets, social computing, and socio-economic and communications networks.