Neural Information Processing Systems http://nips.cc/

Inference in the introduced linear Gaussian SSM is straightforward and can be performed using Kalman prediction and observation updates.

Forboundednonconvex losses and a batch sizem = 1, we additionally show that both generalization error and learning rate are independent ofd and K, and remain essentially the same asfortheSGD, evenfortwofunction evaluations.

This work was accomplished when Liyi Chen was an intern at Alibaba Cloud Computing.

Continual lifelong learning is essential to many applications.

However, in many real-world applications, collecting sufficient exploratory interactions is usually impractical, because online datacollection canbecostlyorevendangerous, suchasinhealthcare [4]andautonomous driving [5]. To address this challenge, offline RL [6, 7] develops a new learning paradigm that trains RL agents only with pre-collected offline datasets and thus can abstract away from the cost of online exploration [8-17].

However, in many real-world applications, collecting sufficient exploratory interactions is usually impractical, because online datacollection canbecostlyorevendangerous, suchasinhealthcare [4]andautonomous driving [5]. To address this challenge, offline RL [6, 7] develops a new learning paradigm that trains RL agents only with pre-collected offline datasets and thus can abstract away from the cost of online exploration [8-17].