reinforcement learning

Intelligent Vehicular Routing and Optimal Resource Allocation Using Deep Reinforcement Learning for Urban VANETs

The proliferation of connected and autonomous vehicles has intensified the demand for robust Vehicular Ad-hoc Networks (VANETs). A critical challenge in urban VANETs is the co-design of efficient routing protocols and optimal communication resource allocation under highly dynamic and resource-constrained conditions. Tr...

Data Center Thermal Optimization Using Graph Neural Networks

Data center cooling accounts for a significant fraction of facility energy consumption, motivating the development of intelligent control strategies that improve efficiency while maintaining thermal safety. This study presents a graph-based digital twin for data center thermal optimization that enables safe online rein...

Extinction reduces the generalization of context-dependent behavior

In the Extinction Makes Acquisition Context-Specific (EMACS) effect, new learning acquired during extinction becomes increasingly dependent on the extinction context. The present study examined whether extinction modifies the behavioral generalization of contextual control and whether these changes are accompanied by a...

AlphaZero & MCTS: A Deep Reinforcement Learning Approach to Pokémon TCG

Raven: From Scientific Data to Trustworthy Autonomous Discovery

Transformational AI for science will depend less on isolated model capability than on the quality of the scientific data interfaces surrounding each model. Raven is a state-aware agentic architecture for turning heterogeneous data, simulations, literature, and experiments into auditable discovery decisions. Its central...

Research on the problems of artificial intelligence empowering comprehensive budget management

This paper constructs a guarantee system for the implementation of AI budgeting, while also pointing out the practical obstacles currently faced by intelligent budget transformation, such as insufficient model interpretability, data silos, a shortage of compound financial talent, and imperfect supporting institutions.

Closed-loop non-invasive ventilation in critical care: current evidence and clinical perspectives

Non-invasive ventilation (NIV) is a cornerstone of acute and perioperative respiratory support, but its effectiveness is constrained by patient selection, interface leaks, patient-ventilator asynchrony, excessive inspiratory effort, and delayed recognition of treatment failure. Closed-loop and adaptive approaches may a...

MXene-based interface-engineering electrocatalysts and artificial intelligence-driven strategies for advanced anion exchange membrane water and seawater electrolysis

Green hydrogen production through water electrolysis is a promising pathway for renewable-energy integration and low-carbon development. Among electrolysis technologies, Anion Exchange Membrane Water Electrolysis (AEMWE) is attracting increasing attention because it enables alkaline operation, uses earth-abundant catal...

The Reassessment That Did Not Travel: What the 2020 Arcade Result About Exploration Bonuses Establishes, the Six Conditions That Made It Informative, and Which of Them the Language-Model Revival Drops

(c) 2026 Pranay Mahendrakar. Licensed under CC BY 4.0. Exploration bonuses have returned. Between 2025 and 2026 at least six frameworks added an intrinsic novelty or uncertainty term to reinforcement learning with verifiable rewards for language models, each naming a classical antecedent - prediction error, pseudo-coun...

Data Center Thermal Optimization Using Graph Neural Networks

Data center cooling accounts for a significant fraction of facility energy consumption, motivating the development of intelligent control strategies that improve efficiency while maintaining thermal safety. This study presents a graph-based digital twin for data center thermal optimization that enables safe online rein...

Dynamic resource allocation and energy efficiency optimization for a UAV-assisted ISAC system based on deep reinforcement learning

Simulations show the SAC-based approach outperforms benchmarks in task completion rate, energy consumption, and convergence, and validate the maximum entropy mechanism's effectiveness in enhancing exploration and generalization in UAV-assisted ISAC environments.

THE LOGIC OF LEARNING MACHINES - How Artificial Intelligence Sees, Learns, Reasons, and Creates

The Logic of Learning Machines: How Artificial Intelligence Sees, Learns, Reasons, and Creates is a comprehensive educational textbook and reference work introducing the foundations, methods, applications, limitations, and future directions of artificial intelligence.Written for students and readers who want to progres...