arXiv research

Paper tackles overestimation bias in continuous control, improving performance by 25%.

Proposes hybrid reinforcement learning for both discrete and continuous control problems.

A hybrid model combines Q-learning and PID controller for continuous vehicle control.

Continuous-time MBRL framework tackles control systems with Bayesian ODEs.

Paper tackles risk-sensitive impulse control for continuous-time processes.

Continuous reinforcement learning such as DDPG and A3C are widely used in robot control and autonomous driving. However, both methods have theoretical weaknesses. While DDPG cannot control noises in the control process, A3C does not satisfy the continuity conditions under the Gaussian policy. To address these concerns,…

Automatically finds strong neural network topologies for continuous control tasks.

Study Transformer layers under cross-entropy training using mean field control.

New framework for analyzing games with multi-dimensional singular controls and non-linear jumps.

Efficient deep policy gradient method for continuous-time control problems.

Tabular Q-Learning with learned state abstractions solves continuous control tasks.

We prove continuity of a controlled SDE solution in Skorokhod's $M_1$ and $J_1$ topologies and also uniformly, in probability, as a non-linear functional of the control strategy. The functional comes from a finance problem to model price impact of a large investor in an illiquid market. We show that $M_1$-continuity is…

Neural Laplace Control tackles offline RL for continuous-time delayed systems with irregular observations.

Deep learning solves complex stochastic control with jumps.

New framework for policy gradient methods in continuous time reinforcement learning.

HiDe learns hierarchical control for complex tasks by separating planning and control.

Logarithmic regret achieved in continuous-time linear-quadratic reinforcement learning.

Neural system optimizes glucose levels in diabetics.

Hybrid systems are characterized by having an interaction between continuous dynamics and discrete events. The contribution of this paper is to provide hybrid systems with a novel geometric formulation so that controls can be added. Using this framework we describe some new global controllability tests for hybrid contr…

Motivated by the recent applications of game-theoretical learning techniques to the design of distributed control systems, we study a class of control problems that can be formulated as potential games with continuous action sets, and we propose an actor-critic reinforcement learning algorithm that provably converges t…

Deep RBVFs improve continuous control in RL.

ACSSM models irregular time series with continuous dynamics.

This paper establishes the existence of a unique nonnegative continuous viscosity solution to the HJB equation associated with a Markovian linear-quadratic control problems with singular terminal state constraint and possibly unbounded cost coefficients. The existence result is based on a novel comparison principle for…

Action chunking and data exploration improve behavior cloning in robotics.

A new approach models exploration in continuous-time RL using random measures.

New NN design for nonlinear systems control with guarantees.

Study shows certainty equivalent policy minimizes regret in continuous-time systems.

New method uses differential equations for better counterfactual analysis.

Study shows how to control jump-diffusion processes with stable feedback controls in reinforcement learning.

Continuous control imitation learning fails if expert actions are smooth.

A new framework for generative modeling using controlled vector fields.

Algorithm improves continuous control with context, achieving provable efficiency.

As energy markets begin clearing at sub-hourly rates, their interaction with load control systems becomes a potentially important consideration. A simple model for the control of thermal systems using market-based power distribution strategies is proposed, with particular attention to the behavior and dynamics of elect…

New model learning objective improves continuous control tasks.

This paper develops q-learning methods for mean-field control problems.

We solve continuous-time reinforcement learning using distributional Hamilton-Jacobi-Bellman equations.

Novel Bayesian meta-reinforcement learning framework improves traffic signal control robustness.

Faster policy learning via continuous-time gradients.

The paper designs neural networks with assurance for controlling nonlinear systems.

PFPN uses particle filtering to improve character control in physics-based simulations.

DyNODE uses neural ODEs to model system dynamics in continuous control tasks.

Study uses actor-critic method for continuous-time mean-field control with entropy regularisation.

Computational Drug Repositioning (CDR) is the task of discovering potential new indications for existing drugs by mining large-scale heterogeneous drug-related data sources. Leveraging the patient-level temporal ordering information between numeric physiological measurements and various drug prescriptions provided in E…

Reinforcement learning algorithms rely on exploration to discover new behaviors, which is typically achieved by following a stochastic policy. In continuous control tasks, policies with a Gaussian distribution have been widely adopted. Gaussian exploration however does not result in smooth trajectories that generally c…

Reliable and effective multi-task learning is a prerequisite for the development of robotic agents that can quickly learn to accomplish related, everyday tasks. However, in the reinforcement learning domain, multi-task learning has not exhibited the same level of success as in other domains, such as computer vision. In…

Study policy gradient for large-agent mean-field control and game in continuous time.

Nonlinear optimal control problems are often solved with numerical methods that require knowledge of system's dynamics which may be difficult to infer, and that carry a large computational cost associated with iterative calculations. We present a novel neurobiologically inspired hierarchical learning framework, Reinfor…

We provide a framework for incorporating robustness -- to perturbations in the transition dynamics which we refer to as model misspecification -- into continuous control Reinforcement Learning (RL) algorithms. We specifically focus on incorporating robustness into a state-of-the-art continuous control RL algorithm call…