Optimality, Approximation and Robustness in Auctions

Data dell'evento: 
Martedì, 16 Ottobre, 2018 - 12:00
Room B203, II floor
Yiannis Giannakopoulos
In this talk we will briefly overview some of our recent results in the area of optimal
We will first study a standard Bayesian auction setting, where multiple bidders have
i.i.d. valuations for a single item, showing that for the natural class of Monotone
Hazard Rate (MHR) distributions, offering the same, take-it-or-leave-it price to all
bidders achieves an (asymptotically) optimal revenue. We will then present a general
duality-theory framework for revenue maximization in additive Bayesian auctions
involving many bidders, multiple items and arbitrary joint value distributions. We
will demonstrate the power of the framework by applying it to special single-bidder
settings with independent item valuations drawn from various distributions of
interest, to design both exact and approximately optimal auctions. Previous exact
solutions were essentially only known for up to two items and for a very limited
number of specific distributions. Finally, we will study a dynamic market setting
where an intermediary interacts with an unknown large sequence of agents that can be
either sellers or buyers: their identities, as well as the sequence length, are
decided in an adversarial, online way. The intermediary has some prior,
distributional knowledge of the agents' values for the items, and uses a
posted-price selling mechanisms.
Some of the related papers have appeared in EC'14, ICALP '15 and '17, WINE'18, and
they can be found in the following links:


Stefano Leonardi