reinforcement learning

A data-driven reinforcement learning framework for optimal balance control and stabilization of a rotary inverted pendulum

Adaptive SOH-Aware Frequency Regulation for Energy Storage Systems via Deep Reinforcement Learning

Reinforcement learning optimization of a ⁶Li magneto-optical trap: DDPG versus SAC

We report an experimental implementation of deep reinforcement learning (RL) for optimizing the loading stage of a ⁶Li magneto-optical trap (MOT) in a high-dimensional continuous control space. An off-policy actor-critic agent observes in-situ fluorescence images at 111 ms intervals and updates seven experimental param...

Energy-efficient FPGA acceleration of reinforcement learning algorithms using high-level synthesis

The results show that pipelined fixed-point FPGA architectures are a good solution to speed up the reinforcement learning process in resource-limited embedded systems.

Reinforcement-Learning-Based Energy Management for a Range-Extended Distributed-Drive Tracked Combine Harvester in Hilly Terrain

Farmland in the hilly and mountainous regions of southern China is characterized by complex terrain and highly variable operating loads. Conventional diesel-powered tracked harvesters are constrained by high crop losses, excessive impurity rates, frequent blockages, and low overall energy-use efficiency. Distributed el...

HG-MARL: A Hierarchical Graph-Based Multi-Agent Reinforcement Learning Framework for Adaptive Policy Simulation in Circular Economies

Reinforcement Learning-Based Dueling DQN for Access Point Selection and Handover Management in Mobile Hybrid LiFi/WiFi Networks

H3C-BEACON: hierarchical hybrid heterogeneous control with Bayesian-elite adaptive coalition network for multi-agent reinforcement learning

H3C-BEACON is presented, a unified hierarchical framework for coordination and control in cooperative MARL that jointly integrates communication, probabilistic belief inference, adaptive coalition formation, and policy stabilisation and significantly improves reproducibility across random initialisations.

Robust Multi-Objective Reinforcement Learning Improves Joint Predicted-Potency and Drug-Likeness Yield in EGFR Molecule Generation

Optimizing a molecular generator against a single activity-model estimate can reward structures for which that estimate is unreliable. We extended our published ChemBERTa–reinforcement-learning (RL) workflow for human epidermal growth factor receptor (EGFR) with scaffold-separated activity modeling, empirically calibra...

Supporting Continuous Improvement in Healthcare with Machine Learning

This project investigates the informatics architecture required to realize a continuous Learning Health System (LHS) in intensive care using offline reinforcement learning. Using mixed shock hemodynamic management as a focal use case, this work aligns organizational needs, algorithmic constraints, and end-user workflow...

Multi-agent meta-reinforcement learning based on contrastive predictive coding—For rapid adaptive traffic control after emergencies

A multi-agent meta-reinforcement learning framework is developed using Contrastive Predictive Coding for traffic state representation that maintained emergency vehicle priority while keeping the effect on regular traffic within a comparatively limited range under the evaluated emergency conditions.

Politeness toward artificial intelligence and the risks of anthropomorphic bias

Against the prevalent social norm of treating conversational artificial intelligence with politeness, this paper explores the cognitive risks brought by human habitual politeness toward chatbots such as ChatGPT. Drawing on historical analogies, virtue ethics and the mechanism of Reinforcement Learning from Human Feedba...