Novel frameworks for auctions and optimization

Abstract

This thesis contains two parts. Part I introduces novel frameworks for modeling uncertainty in auctions. This enables us to provide robust analysis to alternative specifications of preferences and information structures in Vickrey and VCG auctions. Part II introduces novel frameworks for understanding first-order methods in optimization. This enables us to (1) break 20-year barriers on the running time used for solving positive linear programs, (2) reduce the complexity for solving positive semidefinite programs, and (3) strengthen the theory of matrix multiplicative weight updates and improve the theory of linear-sized spectral sparsification.