Real-time Learning and Planning in Environments with Swarms: A Hierarchical and a Parameter-based Simulation Approach - Dr Leandro Soriano Marcolino - Palestra do Grupo de Pesquisa de Middleware

No dia 25 de setembro, sexta-feira, às 10h, no canal do INF no YouTube, será transmitida a palestra "Real-time Learning and Planning in Environments with Swarms: A Hierarchical and a Parameter-based Simulation Approach", com o Prof. Dr. Leandro Soriano Marcolino, da Lancaster University.

A palestra será em português e faz parte das reuniões do Grupo de Pesquisa de Middleware, do Instituto de Informática da UFG.

Título: Real-time Learning and Planning in Environments with Swarms: A Hierarchical and a Parameter-based Simulation Approach

Swarms can be applied in many relevant domains, such as patrolling or rescue. They usually follow simple local rules, leading to complex emergent behavior. Given their wide applicability, an agent may need to take decisions in an environment containing a swarm that is not under its control, and that may even be an antagonist. Predicting the behavior of each swarm member is a great challenge, and must be done under real time constraints, since they usually move constantly following quick reactive algorithms. We propose the first two solutions for this novel problem, showing integrated on-line learning and planning for decision-making with unknown swarms: (i) we learn an ellipse abstraction of the swarm based on statistical models, and predict its future parameters using time-series; (ii) we learn algorithm parameters followed by each swarm member, in order to directly simulate them. We find in our experiments that we are significantly faster to reach an objective than local repulsive forces, at the cost of success rate in some situations. Additionally, we show that this is a challenging problem for reinforcement learning.

Bio:
Leandro Soriano Marcolino is an assistant professor at Lancaster
University. He obtained his doctorate degree at University of Southern
California (USC), advised by Milind Tambe. He has published in several
key conferences in AI, robotics and machine learning, such as AAAI,
AAMAS, IJCAI, NIPS, ICRA and IROS. He received the best dissertation and
the best research assistant award from the Computer Science Department
at USC, had a paper nominated for best paper from the leading
multi-agent conference AAMAS, and had his undergraduate work selected as
the best in the nation by the Brazilian Computer Science Society.
Leandro completed his masters degree in Japan (with the
highly-competitive Monbukagakusho scholarship) and his undergraduate
degree in Brazil at Universidade Federal de Minas Gerais (receiving a
gold medal for finishing the course with the highest grades). His
research interests are multi-agent teamwork, machine learning, robotics;
with emphasis on coordination and collaboration. Leandro has conducted
this research in the context of a variety of domains, such as swarm
robotics, computer Go, social networks, bioinformatics and architectural
design.