This article describes the learning and control capabilities of a biologically constrained bottom-up model of the mammalian cerebellum. Results are presented from six tasks - eyelid conditioning, pendulum balancing, PID control, robot balancing, pattern recognition, and MNIST handwritten digit recognition. These tasks span several paradigms of machine learning including supervised learning, reinforcement learning, control, and pattern recognition. Results over these six domains indicate that cerebellar simulation is capable of robustly identifying static input patterns even when randomized across the sensory apparatus. This capability allows the simulated cerebellum to perform several different supervised learning and control tasks. On the other hand, reinforcement learning and temporal pattern recognition both prove problematic due to the delayed nature of error signals and the simulator's inability to solve the credit assignment problem. These results are consistent with previous findings which hypothesize that in the human brain, the basal ganglia is responsible for reinforcement learning while the cerebellum handles supervised learning.

PDF, HTML

IEEE Transactions on Neural Networks and Learning Systems (2016).

@article{TNNLS15-hausknecht, title={Machine Learning Capabilities of a Simulated Cerebellum}, author={Matthew Hausknecht and Wen-Ke Li and Michael Mauk and Peter Stone}, journal={IEEE Transactions on Neural Networks and Learning Systems}, month={January}, url="http://www.cs.utexas.edu/users/ai-lab?hausknecht:tnnls16", year={2016} }