|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | Role of Human- Mediated Dispersal in the Spread of the Pinewood Nematode in China显示文摘 | Christelle Robinet Alain Roques Hongyang Pan | 2009 | PLoS ONE2009，4，2： | 1 |
| 2 | Deposition Effects of Electrostatic and NonElectrostatic Aerial Spray显示文摘In this study,we compared the aerial electrostatic spray with non-electrostatic spray by helicopter R44 and fixed-wing aircraft Y5B equipped with an own-made aerial electrostatic spraying system.The results showed that the spray deposits on targets and leaves of tree crowns were notable enhanced comparing with the non-electrostatic and conventional aerial spray.The spray deposits were fine,uniform and the densities were 1.0-1.75 times higher than the non-electrostatic spray.Deposit rate of the electrostatic spray on the backs of leaves by helicopter R44 was 2.15 times of the non-electrostatic, and 3.31 times of the conventional aerial spray.The deposit rate of the electrostatic spray on the backs of leaves by the Y5B was 2.38 times of the non-electrostatic,and 3.44 times of the conventional aerial spray. | Shu Chaoran Pan Hongyang Zhou Hongping Zhan Min Mei Aihua Ru Yu Jia Zhicheng Chen Guofa | 2012 | Chinese Forestry Science and Technology2012，11，3： | 0 |
| 3 | Adversarial Neural Collaborative Filtering with Embedding Dimension Correlations显示文摘Recently,convolutional neural networks(CNNs)have achieved excellent performance for the recommendation system by extracting deep features and building collaborative filtering models.However,CNNs have been verified susceptible to adversarial examples.This is because adversarial samples are subtle non-random disturbances,which indicates that machine learning models produce incorrect outputs.Therefore,we propose a novel model of Adversarial Neural Collaborative Filtering with Embedding Dimension Correlations,named ANCF in short,to address the adversarial problem of CNN-based recommendation system.In particular,the proposed ANCF model adopts the matrix factorization to train the adversarial personalized ranking in the prediction layer.This is because matrix factorization supposes that the linear interaction of the latent factors,which are captured between the user and the item,can describe the observable feedback,thus the proposed ANCF model can learn more complicated representation of their latent factors to improve the performance of recommendation.In addition,the ANCF model utilizes the outer product instead of the inner product or concatenation to learn explicitly pairwise embedding dimensional correlations and obtain the interaction map from which CNNs can utilize its strengths to learn high-order correlations.As a result,the proposed ANCF model can improve the robustness performance by the adversarial personalized ranking,and obtain more information by encoding correlations between different embedding layers.Experimental results carried out on three public datasets demonstrate that the ANCF model outperforms other existing recommendation models. | Yi Gao Jianxia Chen Liang Xiao Hongyang Wang Liwei Pan Xuan Wen Zhiwei Ye Xinyun Wu | 2023 | Data Intelligence2023，5，3： | 0 |
| 4 | Model architecture level privacy leakage in neural networks显示文摘Privacy leakage is one of the most critical issues in machine learning and has attracted growing interest for tasks such as demonstrating potential threats in model attacks and creating model defenses.In recent years,numerous studies have revealed various privacy leakage risks(e.g.,data reconstruction attack,membership inference attack,backdoor attack,and adversarial attack)and several targeted defense approaches(e.g.,data denoising,differential privacy,and data encryption).However,existing solutions generally focus on model parameter levels to disclose(or repair)privacy threats during the model training and/or model interference process,which are rarely applied at the model architecture level.Thus,in this paper,we aim to exploit the potential privacy leakage at the model architecture level through a pioneer study on neural architecture search(NAS)paradigms which serves as a powerful tool to automate a neural network design.By investigating the NAS procedure,we discover two attack threats in the model architecture level called the architectural dataset reconstruction attack and the architectural membership inference attack.Our theoretical analysis and experimental evaluation reveal that an attacker may leverage the output architecture of an ongoing NAS paradigm to reconstruct its original training set,or accurately infer the memberships of its training set simply from the model architecture.In this work,we also propose several defense approaches related to these model architecture attacks.We hope our work can highlight the need for greater attention to privacy protection in model architecture levels(e.g.,NAS paradigms). | Yan LI Hongyang YAN Teng HUANG Zijie PAN Jiewei LAI Xiaoxue ZHANG Kongyang CHEN Jin LI | 2024 | Science China(Information Sciences)2024，67，3： | 0 |
| 5 | AIFM1 variants associated with auditory neuropathy spectrum disorder cause apoptosis due to impaired apoptosis-inducing factor dimerization显示文摘Auditory neuropathy spectrum disorder(ANSD)represents a variety of sensorineural deafness conditions characterized by abnormal inner hair cells and/or auditory nerve function,but with the preservation of outer hair cell function.ANSD represents up to 15%of individuals with hearing impairments.Through mutation screening,bioinformatic analysis and expression studies,we have previously identified several apoptosis-inducing factor(AIF)mitochondria-associated 1(AIFM1)variants in ANSD families and in some other sporadic cases.Here,to elucidate the pathogenic mechanisms underlying each AIFM1 variant,we generated AIF-null cells using the clustered regularly interspersed short palindromic repeats(CRISPR)/CRISPR-associated protein 9(Cas9)system and constructed AIF-wild type(WT)and AIF-mutant(mut)(p.T260A,p.R422W,and p.R451Q)stable transfection cell lines.We then analyzed AIF structure,coenzyme-binding affinity,apoptosis,and other aspects.Results revealed that these variants resulted in impaired dimerization,compromising AIF function.The reduction reaction of AIF variants had proceeded slower than that of AIF-WT.The average levels of AIF dimerization in AIF variant cells were only 34.5%-49.7%of that of AIF-WT cells,resulting in caspase-independent apoptosis.The average percentage of apoptotic cells in the variants was 12.3%-17.9%,which was significantly higher than that(6.9%-7.4%)in controls.However,nicotinamide adenine dinucleotide(NADH)treatment promoted the reduction of apoptosis by rescuing AIF dimerization in AIF variant cells.Our findings show that the impairment of AIF dimerization by AIFM1 variants causes apoptosis contributing to ANSD,and introduce NADH as a potential drug for ANSD treatment.Our results help elucidate the mechanisms of ANSD and may lead to the provision of novel therapies. | YUE QIU HONGYANG WANG HUAYE PAN JING GUAN LEI YAN MINGJIE FAN HUI ZHOU XUANHAO ZHOU KAIWEN WU ZEXIAO JIA QIANQIAN ZHUANG ZHAOYING LEI MENGYAO LI XUE DING AIFU LIN YONG FU DONG ZHANG QIUJU WANG QINGFENG YAN | 2023 | Journal of Zhejiang University-Science B(Biomedicine & Biotechnology)2023，24，2： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员