Palestras e Seminários
Salvar atividade no Google Calendar C4AI


Deep learning has become the method of choice for tasks related to perception and natural language processing. Yet, it remains an open question as to how this breakthrough can extend to tasks typically seen as requiring reasoning, such as planning. In this talk, I will introduce neural network architectures that leverage the structure of planning problems, as captured by traditional planning representations, to learn generalized policies and heuristic estimators with minimal training time and data. The obtained policies can adapt to planning problems with different initial states, goals, objects, and sizes. Furthermore, the learnt heuristics generalize to problems from different planning domains. While there are significant limitations, these architectures have showcased remarkable performance on select planning competition benchmark domains. I will conclude by offering a glimpse into my ongoing work in this area.

Brief bio:
Dr. Felipe Trevizan is a Senior Lecturer at the School of Computing at the Australian National University (ANU). Previously, Felipe was a Senior Research Scientist at NICTA (now Data61/CSIRO). Felipe earned his Ph.D. (2013) and M.Sc. (2010) in Machine Learning from Carnegie Mellon University and his M.Sc. (2006) and B.Sc. (2004) in Computer Science from University of São Paulo. Felipe’s research interests lie at the intersection of Artificial Intelligence, Operations Research and Machine Learning including automated planning and scheduling, reasoning under uncertainty, heuristic search, and learning for planning. Along with colleagues and students, Felipe is the co-recipient of the 2016 best paper award from the Transport Research Board and the best paper award at the International Conference on Automated Planning and Scheduling (ICAPS) in 2016 and 2017.

About Dr. Dr. Felipe Trevizan – WebSite:   

#c4ai #ArtificialIntelligence #Machine Learning #Planning #PerspectivesInAI



Mais informações:
site do evento:


© 2023 Instituto de Ciências Matemáticas e de Computação