C Yildiz, M Heinonen… - … on Machine Learning, 2021 - proceedings.mlr.press
… To contrast the robustness of discrete and continuoustime techniques, we evaluate MPETS and ENODE on more realistic datasets of irregularly sampled and noisy data sequences, …
H Wang, T Zariphopoulou, XY Zhou - Journal of Machine Learning …, 2020 - jmlr.org
… We consider reinforcementlearning (RL) in continuoustime with continuous feature and … formulation for the feature dynamics that captures learning under exploration, with the resulting …
LC Baird - Proceedings of 1994 IEEE International Conference …, 1994 - ieeexplore.ieee.org
… is applicable to reinforcementlearning systems working in continuoustime (or discrete time with small time steps) for which standard algorithms such as Q-learning are not applicable. …
S Bradtke, M Duff - Advances in neural information …, 1994 - proceedings.neurips.cc
… ) - extending the domain of applicability to continuoustime. This effort was originally motivated by the desire to apply reinforcementlearning methods to problems of adaptive control of …
BA Wallace, J Si - … on Neural Networks and Learning Systems, 2023 - ieeexplore.ieee.org
This exposition discusses continuous-timereinforcementlearning (CT-RL) for the control of affine nonlinear systems. We review four seminal methods that are the centerpieces of the …
A Perrusquía, W Yu - International Journal of Systems Science, 2021 - Taylor & Francis
… ) are unknown, we can use reinforcementlearning to obtain an optimal and robust control … (15) The above expression can be written as the integral reinforcementlearning (IRL) form (16…
M García-Galicia, AA Carsteanu… - Expert Systems with …, 2019 - Elsevier
… of policy optimization in the context of continuous-timeReinforcementLearning (RL), a branch … The underlying asset portfolio process is assumed to possess a continuous-time discrete-…
… This paper generalizes the MAXQ method to continuous-time discounted and … reinforcement learning algorithms: continuous-time discounted reward MAXQ and continuous-time …
… continuoustime. Next, we show how spiking neurons can implement a critic, to represent and learn … Third, we discuss a spiking neuron actor, and how it can represent and learn a policy…