Search

Your search keyword '"Xu, Wenlu"' showing total 17 results
Start Over Author "Xu, Wenlu"
17 results on '"Xu, Wenlu"'

2. Byzantine-Resilient Federated Learning at Edge

Author

Tao, Youming, Cui, Sijia, Xu, Wenlu, Yin, Haofei, Yu, Dongxiao, Liang, Weifa, and Cheng, Xiuzhen

Subjects
Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Artificial Intelligence
Abstract

Both Byzantine resilience and communication efficiency have attracted tremendous attention recently for their significance in edge federated learning. However, most existing algorithms may fail when dealing with real-world irregular data that behaves in a heavy-tailed manner. To address this issue, we study the stochastic convex and non-convex optimization problem for federated learning at edge and show how to handle heavy-tailed data while retaining the Byzantine resilience, communication efficiency and the optimal statistical error rates simultaneously. Specifically, we first present a Byzantine-resilient distributed gradient descent algorithm that can handle the heavy-tailed data and meanwhile converge under the standard assumptions. To reduce the communication overhead, we further propose another algorithm that incorporates gradient compression techniques to save communication costs during the learning process. Theoretical analysis shows that our algorithms achieve order-optimal statistical error rate in presence of Byzantine devices. Finally, we conduct extensive experiments on both synthetic and real-world datasets to verify the efficacy of our algorithms.

Published
2023

3. Incentive-Aware Recommender Systems in Two-Sided Markets

Author

Dai, Xiaowu, Xu, Wenlu, Qi, Yuan, and Jordan, Michael I.

Subjects
Computer Science - Information Retrieval, Computer Science - Machine Learning, Statistics - Machine Learning
Abstract

Online platforms in the Internet Economy commonly incorporate recommender systems that recommend products (or "arms") to users (or "agents"). A key challenge in this domain arises from myopic agents who are naturally incentivized to exploit by choosing the optimal arm based on current information, rather than exploring various alternatives to gather information that benefits the collective. We propose a novel recommender system that aligns with agents' incentives while achieving asymptotically optimal performance, as measured by regret in repeated interactions. Our framework models this incentive-aware system as a multi-agent bandit problem in two-sided markets, where the interactions of agents and arms are facilitated by recommender systems on online platforms. This model incorporates incentive constraints induced by agents' opportunity costs. In scenarios where opportunity costs are known to the platform, we show the existence of an incentive-compatible recommendation algorithm. This algorithm pools recommendations between a genuinely good arm and an unknown arm using a randomized and adaptive strategy. Moreover, when these opportunity costs are unknown, we introduce an algorithm that randomly pools recommendations across all arms, utilizing the cumulative loss from each arm as feedback for strategic exploration. We demonstrate that both algorithms satisfy an ex-post fairness criterion, which protects agents from over-exploitation. All code for using the proposed algorithms and reproducing results is made available on GitHub.

Published
2022

5. Association between a single‐nucleotide polymorphism of the angiotensin‐converting enzyme gene and susceptibility to systemic lupus erythematosus in the Hainanese population of China.

Author

Zhang, Qi, Zhuang, Yanping, Zhang, Xuan, Lin, Guiling, Wu, Huitao, He, Ziman, Wang, Zhe, Xu, Wenlu, Yin, Xiyu, Su, Linglan, Jia, Xiaokang, and Gong, Aimin

Subjects
GENETIC models, ANGIOTENSIN converting enzyme, SYSTEMIC lupus erythematosus, DISTRIBUTION (Probability theory), HAPLOTYPES, SINGLE nucleotide polymorphisms
Abstract

The angiotensin‐converting enzyme (ACE) gene plays a significant role in regulating immune responses and inflammatory processes, thus impacting the susceptibility to systemic lupus erythematosus (SLE). Understanding how single‐nucleotide polymorphisms (SNPs) within the ACE gene contribute to the genetic susceptibility to SLE is essential for comprehending the disease's aetiology. Therefore, exploring this relationship in the Hainan region of China is crucial for gaining insights into the pathogenesis of SLE. This study comprised 428 participants, including 214 SLE patients and 214 healthy controls. Clinical data were gathered, and blood samples were collected. Genotyping of three SNPs (rs4459609, rs4309, rs1987692) within the ACE gene was performed using SNaPshot technology. The frequencies of alleles and genotypes of these three SNPs were compared between the SLE and control groups. Combining different genetic models and haplotype analysis, the correlation between ACE gene polymorphisms and SLE was investigated. Both study groups exhibited conformity with the Hardy–Weinberg genetic equilibrium (p >.05). Significant differences were observed in the genotype frequency distributions of ACE genes rs4459609, rs4309 and rs1987692 between the SLE and control groups (p =.009,.008,.032, respectively). The frequency of allele T at rs4309 was significantly higher in the SLE group than in the control group, correlating significantly with increased SLE risk (odds ratio [OR] = 1.527, 95% confidence interval [CI] = 1.147–2.035). Associations among ACE rs4459609, rs4309 and rs1987692 polymorphisms and increased susceptibility to SLE were found under co‐dominant and dominant models (p <.05, with OR values and 95% CI greater than 1). Linkage disequilibrium was observed among rs4459609, rs4309 and rs1987692, and haplotype analysis revealed a significantly higher frequency of the CCA haplotype in the control group compared to the SLE group (p <.001). The ACA and ATA haplotypes showed significantly higher frequencies in the SLE group than in the control group (p =.014, p =.013, respectively). ACE gene polymorphisms are associated with the genetic susceptibility to SLE. The AC and AA genotypes at the rs4459609 locus, the TT genotype and T allele at the rs4309 locus and the AC and CC genotypes at the rs1987692 locus may serve as risk factors for the development of SLE. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

17. Exploring the Relationship between ACE Gene Variants and Systemic Lupus Erythematosus Susceptibility in the Hainan Population through Genetic Association Analysis.

Author

Lin G, Zhuang Y, Zhang X, Zhang Q, Wu H, Xu W, Wang Z, He Z, Su L, Jia X, and Gong A

Abstract

Background: The purpose of this study was to determine the association between single nucleotide polymorphisms (SNPs) at the rs4331, rs4341, and rs4351 loci of the angiotensinconverting enzyme (ACE) gene and genetic susceptibility to systemic lupus erythematosus (SLE) in the Hainan population., Methods: This study involved a total of 428 participants, with 214 individuals diagnosed with SLE and an equal number of healthy controls. The SNaPshot sequencing technique was used to determine the base sequences at the ACE gene rs4331, rs4341, and rs4351 loci in the study subjects. Logistic regression was employed to compare the frequency distribution of genotypes and allele frequencies at each locus between the case group and the control group. HaploView 4.2 software was used to analyze the relationship between haplotypes at each locus and genetic susceptibility to SLE., Results: The GG genotype and G allele frequency at the rs4341 locus were higher in the case group compared to the control group. In the rs4341 recessive model, carriers of the GG genotype were more likely to develop SLE compared to carriers of the CG+CC genotype (OR = 1.889, 95% CI: 1.195-2.988, P = 0.006). In the rs4351 overdominant model, carriers of the AC genotype had an increased risk of developing SLE compared to carriers of the AA+CC genotype (OR = 1.514, 95% CI: 1.033-2.219, P = 0.033). The rs4341 and rs4351 loci exhibited linkage disequilibrium, and the CA haplotype (OR = 0.630, 95% CI: 0.481-0.826, P = 0.001) was a protective factor against SLE. The GA haplotype (OR = 2.849, 95% CI: 1.901-4.270, P < 0.01) and the CC haplotype (OR = 2.309, 95% CI: 1.210-4.405, P = 0.009) were risk factors for genetic susceptibility to SLE in the Hainan population., Conclusion: The rs4341 locus of the ACE gene is associated with genetic susceptibility to SLE in the Hainan population., (Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.net.)

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results