You are here

Optimality, Constraints and Feasibility: A Complex Equilibrium In A Dynamic World

Alessandra Sala, Nokia Bell Labs
3-4pm  14th Feb 2018

Abstract

The digital era comes with the promise of services available to us anywhere anytime but it also brings growing challenges in controlling and managing these stochastic, heterogeneous and dynamic digital environments. The complexity and the dynamicity of these new environments make it more and more difficult to apply standard optimization approaches based on a priori specification, formal system models and learnable profiles of the system behaviors.

A paradigm of continuous learning via interactions brings strong promises to realize adaptive control mechanisms. By starting with a little or no knowledge about system, the control agents takes actions and learns their efficiency through the quality of the observed feedback from the environment. This approach allows to hide the inherent complexity of the environment and to adapt dynamically to its changing conditions.

This seminar will overview some examples of practical optimization problems and then review the basic ideas of learning via interactions in uncertain environments. i.e. Reinforcement Learning. Finally, it will conclude with a use case that leverages a reinforcement learning model to solve an optimization problem and some future direction in the space.

Short Bio

Alessandra Sala leads the Analytics Research in Nokia Bell-Labs and is responsible to deliver breakthrough research assets to create new market opportunities and technology that has the potential to change our human lives. In her prior appointment, she was the technical manager of the Data Analytics and Operations Research group in Bell Labs Ireland. Before that, she held a research associate position in the Department of Computer Science at University of California Santa Barbara. During this appointment, she was a key contributor of several funded proposals from National Science Foundation in USA and her research was awarded with the Cisco Research Award in 2011. Before UCSB, she completed her Ph.D in Computer Science at University of Salerno, Italy.

Her research focus lies on distributed algorithms, data analytics and complexity analysis with an emphasis on graph algorithms and recently AI, machine learning and deep learning. In her previous research she has developed efficient distributed routing algorithms that support robust and flexible application level services such as scalable search, flexible data dissemination, and reliable anonymous communication.

She was recently the Technical Program Committee Chair of WWW 2016, the general chair of ACM COSN 2014 and has served on the TPC of several networking and data mining conferences as IEEE INFOCOM, WWW, P2P etc. Two of her recent awards are: Distinguished Member of the 2015 IEEE INFOCOM Technical Program Committee and Institute of Telecommunications Professionals (ITP) Innovator of the Year 2017.

Venue

Lloyd LB08