Function Tables for Secure Distributed Matrix Multiplication
该研究通过引入函数表来解决安全分布式矩阵乘法中的隐私和解码问题，提出了一种新的线性方案，并确定了在不同条件下的最优工作节点数量。
0 citationsRead paper
Selected work
Representative Papers
Conformal Robust Set Estimation
Standard conformal prediction struggles to guarantee reliable coverage under outliers or heavy-tailed distributions. This work proposes a novel nonconformity scoring method based on the half-sample radius—the distance to the (⌊n/2⌋+1)-th nearest neighbor—thereby introducing geometric robustness into the conformal prediction framework for the first time. The method satisfies marginal validity in finite samples and converges at an exponential rate to the population center set defined by a distance-based functional. Rigorous theoretical analysis yields sharp tail deviation bounds, ensuring both theoretical guarantees and practical robustness for heavy-tailed or multimodal distributions.
0 citationsRead paper
Bounds for the Permutation Flowshop Scheduling Problem: New Framework and Theoretical Insights
To address the insufficient accuracy of upper and lower bound estimates for minimizing makespan in the Permutation Flowshop Scheduling Problem (PFSP), this paper proposes a general extremal framework based on matrix modeling and path-set optimization. The framework uniformly represents both bounds as min-max expressions over paths, yielding a tight lower bound algorithm solvable in polynomial time. Evaluated on the Taillard (120 instances) and VRF (480 instances) benchmark sets, the method improves the best-known lower bounds for 112 and 430 instances, respectively, significantly enhancing bound quality. Theoretically, it advances understanding of PFSP’s asymptotic behavior and provides a novel analytical tool for classical ordering conjectures. The approach combines mathematical rigor, computational efficiency, and strong scalability—demonstrating both practical effectiveness and theoretical depth in combinatorial scheduling optimization.
0 citationsRead paper
Demographic Biases and Gaps in the Perception of Sexism in Large Language Models
This study systematically evaluates the capability of large language models (LLMs) to detect sexist content on the EXIST 2024 Twitter dataset and, for the first time, introduces a multi-annotator demographic lens—e.g., age and gender—to analyze group-level biases in model outputs. Method: We conduct fine-grained detection experiments across multiple state-of-the-art LLMs and employ statistical modeling to quantify how demographic attributes influence inter-annotator agreement between models and human annotators. Results: While LLMs demonstrate baseline discriminatory text detection ability, they fail to replicate the diversity of human judgments on sexism—exhibiting systematic disparities across age and gender subgroups. Contribution: Our work exposes a fundamental limitation in current LLM fairness modeling: the absence of sociodemographically grounded calibration. We propose integrating pluralistic social perspectives into model development as a necessary pathway toward equitable AI, offering empirical evidence and methodological guidance for building more representative, ethically robust AI systems.
0 citationsRead paper
Optimal Experimental Design for Microplastics Sampling Experiments
Microplastic pollution monitoring faces statistical and logistical challenges in jointly optimizing sampling scale and laboratory analysis depth under resource constraints. Method: This paper proposes a Bayesian optimal experimental design framework that employs a conjugate Poisson–multinomial prior model to integrate prior knowledge and quantify uncertainty; it defines information gain via posterior variance minimization and incorporates realistic cost constraints to jointly optimize the number of spatial sampling locations and the analytical depth (i.e., particle composition characterization) per location. Contribution/Results: Compared with conventional fixed-allocation strategies, our approach significantly improves posterior estimation accuracy and decision robustness on both synthetic and field-collected data. It delivers an interpretable, principled resource allocation scheme that balances data quality and research efficiency, thereby establishing a novel paradigm for standardized and intelligent environmental microplastic monitoring.
0 citationsRead paper
Recent publications
Latest Papers
Function Tables for Secure Distributed Matrix Multiplication
该研究通过引入函数表来解决安全分布式矩阵乘法中的隐私和解码问题，提出了一种新的线性方案，并确定了在不同条件下的最优工作节点数量。
0 citationsRead paper
Conformal Robust Set Estimation
Standard conformal prediction struggles to guarantee reliable coverage under outliers or heavy-tailed distributions. This work proposes a novel nonconformity scoring method based on the half-sample radius—the distance to the (⌊n/2⌋+1)-th nearest neighbor—thereby introducing geometric robustness into the conformal prediction framework for the first time. The method satisfies marginal validity in finite samples and converges at an exponential rate to the population center set defined by a distance-based functional. Rigorous theoretical analysis yields sharp tail deviation bounds, ensuring both theoretical guarantees and practical robustness for heavy-tailed or multimodal distributions.
0 citationsRead paper
Bounds for the Permutation Flowshop Scheduling Problem: New Framework and Theoretical Insights
To address the insufficient accuracy of upper and lower bound estimates for minimizing makespan in the Permutation Flowshop Scheduling Problem (PFSP), this paper proposes a general extremal framework based on matrix modeling and path-set optimization. The framework uniformly represents both bounds as min-max expressions over paths, yielding a tight lower bound algorithm solvable in polynomial time. Evaluated on the Taillard (120 instances) and VRF (480 instances) benchmark sets, the method improves the best-known lower bounds for 112 and 430 instances, respectively, significantly enhancing bound quality. Theoretically, it advances understanding of PFSP’s asymptotic behavior and provides a novel analytical tool for classical ordering conjectures. The approach combines mathematical rigor, computational efficiency, and strong scalability—demonstrating both practical effectiveness and theoretical depth in combinatorial scheduling optimization.
0 citationsRead paper
Demographic Biases and Gaps in the Perception of Sexism in Large Language Models
This study systematically evaluates the capability of large language models (LLMs) to detect sexist content on the EXIST 2024 Twitter dataset and, for the first time, introduces a multi-annotator demographic lens—e.g., age and gender—to analyze group-level biases in model outputs. Method: We conduct fine-grained detection experiments across multiple state-of-the-art LLMs and employ statistical modeling to quantify how demographic attributes influence inter-annotator agreement between models and human annotators. Results: While LLMs demonstrate baseline discriminatory text detection ability, they fail to replicate the diversity of human judgments on sexism—exhibiting systematic disparities across age and gender subgroups. Contribution: Our work exposes a fundamental limitation in current LLM fairness modeling: the absence of sociodemographically grounded calibration. We propose integrating pluralistic social perspectives into model development as a necessary pathway toward equitable AI, offering empirical evidence and methodological guidance for building more representative, ethically robust AI systems.
0 citationsRead paper
Optimal Experimental Design for Microplastics Sampling Experiments
Microplastic pollution monitoring faces statistical and logistical challenges in jointly optimizing sampling scale and laboratory analysis depth under resource constraints. Method: This paper proposes a Bayesian optimal experimental design framework that employs a conjugate Poisson–multinomial prior model to integrate prior knowledge and quantify uncertainty; it defines information gain via posterior variance minimization and incorporates realistic cost constraints to jointly optimize the number of spatial sampling locations and the analytical depth (i.e., particle composition characterization) per location. Contribution/Results: Compared with conventional fixed-allocation strategies, our approach significantly improves posterior estimation accuracy and decision robustness on both synthetic and field-collected data. It delivers an interpretable, principled resource allocation scheme that balances data quality and research efficiency, thereby establishing a novel paradigm for standardized and intelligent environmental microplastic monitoring.
0 citationsRead paper