Emergence of cooperation: A reputation-modulated reinforcement learning
研究通过引入基于声誉调整的强化学习方法，在空间囚徒困境游戏中促进合作行为的出现，展示了声誉不仅提供直接激励还重塑了社会信息环境。
0 citationsRead paper
Selected work
Representative Papers
Identifying parameter couplings and uncertainties of mixed-noise stochastic systems via full-covariance Gaussian mixture network
This study addresses the challenge of parameter identification in stochastic systems with mixed noise, where intractable likelihoods hinder conventional estimation. We propose Penn-GMD, a network that maps trajectories to full-covariance Gaussian mixture distributions and employs surjective parameterization with negative log-likelihood minimization to approximate the true likelihood. The core contribution lies in leveraging full covariance matrices to explicitly reveal parameter coupling and multimodal structures. This approach not only accurately recovers the underlying likelihood distribution but also naturally diagnoses unidentifiability. Consequently, Penn-GMD effectively resolves persistent difficulties in parameter estimation and uncertainty quantification for complex stochastic systems where traditional methods typically fail, offering a robust framework for handling intractable inference problems in noisy dynamical environments.
0 citationsRead paper
Beyond the Edge of Chaos: Stability-Expressivity Transfer in Reservoir Forecasting
Traditional “edge of chaos” criteria struggle to accurately guide the optimal design of reservoir computing for prediction tasks, thereby limiting performance gains. This work addresses this limitation by analyzing the collective dynamics of teacher-forced reservoirs and reveals that the target dynamics are primarily captured by input-modulated stable Lyapunov modes. Building on this insight, the authors propose a novel “stability–expressivity transition index” to precisely identify the optimal spectral radius. This metric overcomes the heuristic constraints of edge-of-chaos approaches and consistently locates optimal parameters across diverse chaotic and quasiperiodic target systems as well as reservoir architectures with varying symmetries, leading to significantly enhanced autonomous prediction performance.
0 citationsRead paper
Student Classroom Behavior Recognition Based on Improved YOLOv8s
This work proposes ALC-YOLOv8s to address challenges in real-world classroom settings, including dense student targets, numerous small objects, frequent occlusions, and imbalanced behavior categories. Building upon YOLOv8s, the model integrates an SPPF-LSKA module to enhance contextual awareness, introduces CFC-CRB and SFC-G2 structures to refine multi-scale feature fusion, and employs ATFLoss to improve learning on minority classes and hard samples. Experimental results demonstrate that the proposed method outperforms the baseline by 1.8% in mAP50 and 2.1% in mAP50–95, surpassing several state-of-the-art detectors and effectively meeting the demand for high-accuracy behavior recognition in complex classroom environments.
0 citationsRead paper
Possibilistic Computation Tree Logic: Decidability and Complete Axiomatization
This paper investigates the satisfiability problem for Possibilistic Computation Tree Logic (PoCTL)—a novel branching-time temporal logic integrating possibility theory to model uncertain systems. Addressing the absence of decidability results and axiomatic foundations for PoCTL, we introduce the first *possibilistic information extraction* technique, enabling the construction of Hintikka structures satisfying local consistency. We prove that PoCTL satisfiability is decidable in exponential time. Furthermore, we establish the first sound and complete axiomatization for PoCTL. These contributions fill a fundamental gap in the decidability theory of possibilistic branching-time logics and provide both a rigorous logical foundation and algorithmic support for model checking and formal verification of uncertain systems.
0 citationsRead paper
Recent publications
Latest Papers
Emergence of cooperation: A reputation-modulated reinforcement learning
研究通过引入基于声誉调整的强化学习方法，在空间囚徒困境游戏中促进合作行为的出现，展示了声誉不仅提供直接激励还重塑了社会信息环境。
0 citationsRead paper
Identifying parameter couplings and uncertainties of mixed-noise stochastic systems via full-covariance Gaussian mixture network
This study addresses the challenge of parameter identification in stochastic systems with mixed noise, where intractable likelihoods hinder conventional estimation. We propose Penn-GMD, a network that maps trajectories to full-covariance Gaussian mixture distributions and employs surjective parameterization with negative log-likelihood minimization to approximate the true likelihood. The core contribution lies in leveraging full covariance matrices to explicitly reveal parameter coupling and multimodal structures. This approach not only accurately recovers the underlying likelihood distribution but also naturally diagnoses unidentifiability. Consequently, Penn-GMD effectively resolves persistent difficulties in parameter estimation and uncertainty quantification for complex stochastic systems where traditional methods typically fail, offering a robust framework for handling intractable inference problems in noisy dynamical environments.
0 citationsRead paper
Beyond the Edge of Chaos: Stability-Expressivity Transfer in Reservoir Forecasting
Traditional “edge of chaos” criteria struggle to accurately guide the optimal design of reservoir computing for prediction tasks, thereby limiting performance gains. This work addresses this limitation by analyzing the collective dynamics of teacher-forced reservoirs and reveals that the target dynamics are primarily captured by input-modulated stable Lyapunov modes. Building on this insight, the authors propose a novel “stability–expressivity transition index” to precisely identify the optimal spectral radius. This metric overcomes the heuristic constraints of edge-of-chaos approaches and consistently locates optimal parameters across diverse chaotic and quasiperiodic target systems as well as reservoir architectures with varying symmetries, leading to significantly enhanced autonomous prediction performance.
0 citationsRead paper
Student Classroom Behavior Recognition Based on Improved YOLOv8s
This work proposes ALC-YOLOv8s to address challenges in real-world classroom settings, including dense student targets, numerous small objects, frequent occlusions, and imbalanced behavior categories. Building upon YOLOv8s, the model integrates an SPPF-LSKA module to enhance contextual awareness, introduces CFC-CRB and SFC-G2 structures to refine multi-scale feature fusion, and employs ATFLoss to improve learning on minority classes and hard samples. Experimental results demonstrate that the proposed method outperforms the baseline by 1.8% in mAP50 and 2.1% in mAP50–95, surpassing several state-of-the-art detectors and effectively meeting the demand for high-accuracy behavior recognition in complex classroom environments.
0 citationsRead paper
Possibilistic Computation Tree Logic: Decidability and Complete Axiomatization
This paper investigates the satisfiability problem for Possibilistic Computation Tree Logic (PoCTL)—a novel branching-time temporal logic integrating possibility theory to model uncertain systems. Addressing the absence of decidability results and axiomatic foundations for PoCTL, we introduce the first *possibilistic information extraction* technique, enabling the construction of Hintikka structures satisfying local consistency. We prove that PoCTL satisfiability is decidable in exponential time. Furthermore, we establish the first sound and complete axiomatization for PoCTL. These contributions fill a fundamental gap in the decidability theory of possibilistic branching-time logics and provide both a rigorous logical foundation and algorithmic support for model checking and formal verification of uncertain systems.
0 citationsRead paper