Two papers have been accepted for publication in IEEE Transactions on Neural Networks and Learning Systems and Neural Networks with the following titles.
Towards Physical Interaction-based Sequential Mobility Assistance using Latent Generative Model of Movement State
t-Soft Update of Target Network for Deep Reinforcement Learning
These are joint researches with my student, and both methods proposed and applied t-momentum, which is robust against learning noise, by focusing on the exponential moving average hidden in machine learning. One in TNNLS is published as open access, so please freely read it! The preprint of another paper in Neural Networks is here.