reinforcement learning

Quantifying Pilot Operational Deviations Based on Flight Data via Inverse Reinforcement Learning and Conformal Prediction

Threshold-based flight data monitoring (FDM) identifies parameter exceedances but provides limited visibility into gradual operational deviation within nominal operating bounds. This study presents a three-layer framework for continuous operational deviation quantification from quick access recorder (QAR) data. A stand...

Coordinated Optimization Strategy for Load Aggregator in Distribution Network Considering Demand Response and Peak Regulation Incentive

With the increasing proportion of flexible resources such as distributed generation, energy storage and demand response in new power systems, load aggregators, as an important subject connecting users and the market, face the challenges of complex load types, large differences in response capabilities, and high peaking...

Urban Vehicle Route Choice Modeling Based on Multi-Source Trajectory Data and Adversarial Inverse Reinforcement Learning

To address the limitations of traditional route choice models, including their dependence on manually constructed path sets, their difficulty in capturing nonlinear decision preferences, and their insufficient consideration of dynamic congestion factors, this paper proposes an urban vehicle route choice modeling method...

Study record and reproduction code for a survey of evaluation design and disclosure requirements in reinforcement learning-based route guidance

Reproduction material for the survey "Study-Area Boundaries in Reinforcement Learning-Based Route Guidance: A Survey of Evaluation Design and Disclosure Requirements" by Donghoun Lee. The package holds the study record of the 93 reviewed studies, one row per study with the sentence of the research article behind each r...

Deep reinforcement learning-based optimal multi-energy dispatch in a droop-controlled islanded microgrid

Optimized AI‐Driven Intelligent Transportation Systems for Secure Communication and Traffic Management in Post‐5G Networks

ABSTRACT After 5G vehicular networks, an intelligent transportation system with an AI‐based framework encompasses the real‐time traffic management, security, and quality of service for safer communication. This paper proposes a new comprehensive framework for improving vehicular network performance based on state‐of‐th...

Empowering a first-time mother through postnatal education is not just teaching newborn care—it is nurturing confidence, protecting a precious life, and building a healthier future for generations

The first weeks after childbirth are a period of rapid learning, physical recovery, emotional adjustment, and relationship building. For a first-time mother, everyday activities such as feeding, holding, bathing, comforting, protecting warmth, interpreting crying, recognising illness, and deciding when to seek professi...

SeedPro: Place Brachytherapy Seeds Like Expert Clinicians via Hierarchical Reinforcement Agents

Geometric Shielding and Lagrangian Constraints in Safe Reinforcement Learning: A Factorial Study

Safe reinforcement learning requires balancing task performance against safety constraints, yet how different safety mechanisms interact when combined remains an open question. We present a systematic empirical comparison of four reinforcement learning algorithms, PPO, SAC, RCPO (with fixed λ), and Lagrangian PPO (LagP...

Enhancing Pathological VLMs with Cross-Scale Reasoning

StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning

Reinforcement learning is typically treated as a uniform, data-driven optimization process, where updates are guided by rewards and temporal-difference errors without explicitly exploiting global structure. In contrast, dynamic programming methods rely on structured information propagation, enabling efficient and stabl...

The need for uncertainty: why probabilistic LSTMs are key to improving flood predictions and enabling learned warning rules

Code and results accompanying the EGUsphere preprint egusphere-2026-469. This repository implements deep-learning rainfall–runoff models over the CAMELS-CH basins, comparing a deterministic LSTM baseline against three probabilistic output heads — a DRN model (called LD-LSTM in the code), a Bernstein Quantile Network (B...