Talks and Poster Presentations (with Proceedings-Entry):
M. Lechner, R. Grosu, R. Hasani:
"Worm-level Control through Search-based Reinforcement Learning";
Talk: Deep Reinforcement Learning Symposium at the 31st Neural Information Processing Systems (NIPS) Conference, 2017,
Long Beach, CA, USA;
- 2017-12-09; in: "Proceedings of the Deep Reinforcement Learning Symposium at the 31st Neural Information Processing Systems (NIPS) Conference, 2017",
Through natural evolution, nervous systems of organisms formed near-optimal
structures to express behavior. Here, we propose an effective way to create con-
trol agents, by re-purposing the function of biological neural circuit models, to
govern similar real world applications. We model the tap-withdrawal (TW) neural
circuit of the nematode, C. elegans, a circuit responsible for the worm´s reflex-
ive response to external mechanical touch stimulations, and learn its synaptic
and neural parameters as a policy for controlling the inverted pendulum problem.
For reconfiguration of the purpose of the TW neural circuit, we manipulate a
search-based reinforcement learning. We show that our neural policy performs
as good as existing traditional control theory and machine learning approaches.
A video demonstration of the performance of our method can be accessed at
Created from the Publication Database of the Vienna University of Technology.