Kenji Doya
Keynote speakers
- Kenji Doya
- Alon Halevy
- Astrid Prinz
- Andrew Schwartz
- Shankar Subramaniam
- Arthur Toga
Workshop speakers
- Bart ter Haar Romeny
- Uri Eden
- Klaus Linkenkaer-Hansen
- Tim Clark
- Alan Ruttenberg
- Jeffrey Grethe
- Arnd Roth
- Wulfram Gerstner
- Peter Hunter
- Markus Diesmann
- Andrey Semin
- Pietro Liò
- Albert Cardona
- Giorgio Ascoli
- Rolf Kötter
Kenji Doya
Title: Learning algorithms for modeling human behaviors and identifying molecular cascades
Okinawa Institute of Science and Technology, Okinawa, Japan
Abstract: The theoretical frameworks of reinforcement learning and Bayesian inference have played major roles in elucidating the mathematical principles behind action and perception in uncertain environments. For example, the temporal difference (TD) reinforcement learning algorithm brought a breakthrough in understanding the function of the basal ganglia, which have been a dark matter of the brain until the last century. The concepts of Bayesian inference lead to coherent accounts of how we perceive the world from noisy, incomplete sensory inputs and how such precepts are realized by population of stochastically firing neurons. Now these theories and algorithms are not only used for building conceptual models of the brain, but also for developing methods for data-driven modeling and model-based analysis of neurobiological data at different levels, from behaviors to molecules. This keynote reviews recent examples, including characterization of subjects' behaviors by estimated parameters of reinforcement learning and identification of intracellular molecular cascades by Bayesian inference.
Bio sketch: Kenji Doya received a Ph.D. in Mathematical Engineering at Univ. of Tokyo in 1991. He became a research associate at U. Tokyo in 1986, U. San Diego in 1991, and Salk Institute in 1993. He joined ATR in 1994 and became the head of Computational Neurobiology Department, ATR Computational Neuroscience Laboratories in 2003. In 2004, he was appointed as the principal investigator of Neural Computation Unit, Okinawa Institute of Science and Technology. He serves as the co- editor in chief of Neural Networks from 2008 and a board member of International Neural Network Society from 2009. His research interest is in developing robust and flexible learning algorithms and in understanding how the brain can realize them.
LetItB: software tool for Beyasian estimation of parameter distribution of signaling models defined by SBML.