TY - GEN
T1 - Learning to Control Neurons using Aggregated Measurements
AU - Yu, Yao Chi
AU - Narayanan, Vignesh
AU - Ching, Shinung
AU - Li, Jr Shin
N1 - Publisher Copyright:
© 2020 AACC.
PY - 2020/7
Y1 - 2020/7
N2 - Controlling a population of neurons with one or a few control signals is challenging due to the severely underactuated nature of the control system and the inherent nonlinear dynamics of the neurons that are typically unknown. Control strategies that incorporate deep neural networks and machine learning techniques directly use data to learn a sequence of control actions for targeted manipulation of a population of neurons. However, these learning strategies inherently assume that perfect feedback data from each neuron at every sampling instant are available, and do not scale gracefully as the number of neurons in the population increases. As a result, the learning models need to be retrained whenever such a change occurs. In this work, we propose a learning strategy to design a control sequence by using population-level aggregated measurements and incorporate reinforcement learning techniques to find a (bounded, piecewise constant) control policy that fulfills the given control task. We demonstrate the feasibility of the proposed approach using numerical experiments on a finite population of nonlinear dynamical systems and canonical phase models that are widely used in neuroscience.
AB - Controlling a population of neurons with one or a few control signals is challenging due to the severely underactuated nature of the control system and the inherent nonlinear dynamics of the neurons that are typically unknown. Control strategies that incorporate deep neural networks and machine learning techniques directly use data to learn a sequence of control actions for targeted manipulation of a population of neurons. However, these learning strategies inherently assume that perfect feedback data from each neuron at every sampling instant are available, and do not scale gracefully as the number of neurons in the population increases. As a result, the learning models need to be retrained whenever such a change occurs. In this work, we propose a learning strategy to design a control sequence by using population-level aggregated measurements and incorporate reinforcement learning techniques to find a (bounded, piecewise constant) control policy that fulfills the given control task. We demonstrate the feasibility of the proposed approach using numerical experiments on a finite population of nonlinear dynamical systems and canonical phase models that are widely used in neuroscience.
UR - http://www.scopus.com/inward/record.url?scp=85089572803&partnerID=8YFLogxK
U2 - 10.23919/ACC45564.2020.9147426
DO - 10.23919/ACC45564.2020.9147426
M3 - Conference contribution
AN - SCOPUS:85089572803
T3 - Proceedings of the American Control Conference
SP - 4028
EP - 4033
BT - 2020 American Control Conference, ACC 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2020 American Control Conference, ACC 2020
Y2 - 1 July 2020 through 3 July 2020
ER -