Monday, April 24

4193 Etcheverry Hall, 10:30 a.m. - noon

This talk explains how to formulate the now classical problem of optimal liquidation (or optimal trading) inside a Mean Field Game (MFG). This is a noticeable change since usually mathematical frameworks focus on one large trader in front of a " background noise " (or " mean field "). In standard frameworks, the interactions between the large trader and the price are a temporary and a permanent market impact terms, the latter influencing the public price. Here the trader faces the uncertainty of fair price changes too but not only. He has to deal with price changes generated by other similar market participants, impacting the prices permanently too, and acting strategically. Our MFG formulation of this problem belongs to the class of " extended MFG ", we hence provide generic results to address these " MFG of controls ", before solving the one generated by the cost function of optimal trading. We provide a closed form formula of its solution, and address the case of " heterogeneous preferences " (when each participant has a different risk aversion). Last but not least we give conditions under which participants do not need to instantaneously know the state of the whole system, but can " learn " it day after day, observing others' behaviors.

Monday, May 1

3108 Etcheverry Hall, 3:30 p.m. - 5:00 p.m.

Abstract: In this talk, I present and compare two frameworks for distributed-memory branch-and-bound (B&B) tree search. I'll begin with a brief introduction to Stochastic MIPs and to LP relaxation based Branch-and-Bound for solving MIPs. Both of the frameworks presented in this talk are implemented as extensions of PIPS-SBB, which is already a parallel distributed-memory stochastic MIP solver based on the distributed memory stochastic LP solver PIPS-S. As result, both frameworks include two levels of distributed-memory parallelism, for solving LP relaxations and for B&B tree search. The first framework relies on UG, part of the SCIP project, and implements an external coarse parallelization of PIPS-SBB. The second framework, Parallel PIPS-SBB implements a novel internal fine-grained parallelization of PIPS-SBB. We present computational results that evaluate the effectiveness of both frameworks.

Collaborators: Lluis Munguia, Geoffrey Oxberry, Yuji Shinano

Bio: Deepak Rajan is an Operations Research expert in the Center for Applied Scientific Computing (CASC). His research broadly lies in the areas of computational optimization and integer programming, and more specifically in applying such techniques in solving large-scale problems. Recently, Deepak has been working on optimization problems that involve uncertainty in the energy area (power generation, in particular). He has also worked on a variety of optimization problems in other domains, including network design and graph data mining.

Since 2016, Deepak is also an Associate Adjunct Professor of Industrial Engineering and Operations Research at the University of California at Berkeley.


Wednesday, May 3

Shires Hall
2451 Ridge Road
Berkeley, CA 94709

How can we make the world greener? Healthier? More efficient? More joyful? These are big questions, but our students have some creative solutions!

Berkeley's Master of Engineering Program is proud to present this year's Capstone Showcase.
This is an opportunity for the community to learn about the innovative projects that our students have been working on all year long.

Come network with students, faculty, and alumni while trying out the latest innovations in tech!

RSVP to save your spot! Food and drinks will be provided.