Uncertainty quantification via a memristor Bayesian deep neural network for risk-sensitive reinforcement learning

Yudeng Lin, Qingtian Zhang, Bin Gao, Jianshi Tang, Peng Yao, Chongxuan Li, Shiyu Huang, Zhengwu Liu, Ying Zhou, Yuyi Liu, Wenqiang Zhang, Jun Zhu, He Qian, Huaqiang Wu

Abstract: The stochastic features of memristors make them suitable for computation and probabilistic sampling; however, implementing these properties in hardware is extremely challenging. Lin et al. introduce an approach that leverages the cycle-to-cycle read variability of memristors as a physical random variable for in situ, real-time random number generation, and demonstrate it on a risk-sensitive reinforcement learning task.