Abstract: Advertising is a large source of revenue for many internet services. Advertising auctions match buyers and sellers at scale, enabling highly valuable services for their users. An increasingly important aspect of advertising auctions is that of Auto-bidding, which allows advertisers to state their high level goals and ROI constraints, and converts those to optimal per-auction bids. When the constraints are purely monetary budgets, this becomes the well-studied problem of online budgeted allocation. In this talk, we will discuss the formulations of these problems, and present the theory and practical aspects of algorithms in these settings. We will also touch upon some new connections between auctions, bidding, and machine learning.

Bio: Aranyak Mehta is a Distinguished Research Scientist in the Market Algorithms team at Google Research, Mountain View, CA. His research interests lie at the intersection of Algorithms and Economics, with emphasis on the theory and practice of Auction Design, Online Matching, Allocation Algorithms, and the connections to Machine Learning.