期刊文献+
共找到4篇文章
< 1 >
每页显示 20 50 100
基于非线性特征和Cauchy加权M-估计量的鲁棒推荐算法 被引量:5
1
作者 张付志 孙双侠 伊华伟 《计算机学报》 EI CSCD 北大核心 2017年第6期1453-1469,共17页
协同推荐系统作为一种重要的个性化服务模式,在电子商务站点中的应用越来越广泛.然而,各种恶意欺骗和虚假反馈已制约了其应有效能的发挥.如何有效识别和抵御托攻击,确保系统推荐的可信性,是推荐系统研究面临的一大挑战.为了确保推荐的... 协同推荐系统作为一种重要的个性化服务模式,在电子商务站点中的应用越来越广泛.然而,各种恶意欺骗和虚假反馈已制约了其应有效能的发挥.如何有效识别和抵御托攻击,确保系统推荐的可信性,是推荐系统研究面临的一大挑战.为了确保推荐的可信性,人们提出了各种基于矩阵分解的鲁棒协同推荐算法.但是这些推荐算法在面对托攻击时不仅精度损失大,而且鲁棒性较差.为此,该文提出一种基于非线性特征和Cauchy加权M-估计量的鲁棒推荐算法.首先,采用核主成分分析方法提取用户评分矩阵的非线性特征,以充分挖掘推荐系统中用户(或项目)之间的内在关联,最大限度地保留用户和项目的特征信息,提高推荐精度和鲁棒性;然后,引入鲁棒统计中的Cauchy加权M-估计量,并联合矩阵分解模型对用户和项目特征矩阵进行鲁棒参数估计,以限制攻击概貌对参数估计过程产生的影响;最后,设计相应的鲁棒协同推荐算法,并在MovieLens和Netflix数据集上对算法的有效性进行评价.实验结果表明:该文算法在推荐精度和鲁棒性方面明显优于现有的鲁棒推荐算法. 展开更多
关键词 托攻击 鲁棒推荐算法 核函数 Cauchy加权M-估计量 矩阵分解 协同推荐系统 人工智能
在线阅读 下载PDF
A NEW ARTIFICIAL DIFFUSION FACTOR IN THE STREAMLINE UPWIND/PETROV GALERKIN FORMULATION 被引量:1
2
作者 Wei Yong tao, Yu Jian hua, Cao Shu you State Key Hydraulics Laboratory of High Speed Flows, Sichuan University, Chengdu 610065, China 《Journal of Hydrodynamics》 SCIE EI CSCD 2002年第3期76-82,共7页
For the incompressible Navier Stokes equations, a new artificial diffusion factor is put forward in the Streamline Upwind/Petrov Galerkin formulation. The corresponding formulae of finite element methods are derived ... For the incompressible Navier Stokes equations, a new artificial diffusion factor is put forward in the Streamline Upwind/Petrov Galerkin formulation. The corresponding formulae of finite element methods are derived in Newton Raphson form, in which velocity and pressure are iterated synchronously. An element with nine nodes satisfying inf sup condition is established, which has a parabolic velocity interpolation and linear pressure distribution. Four numerical examples are presented, and solutions obtained demonstrate the effectivity of the method proposed. 展开更多
关键词 SU/PG FEM artificial diffusion factor Newton Raphson iteration
原文传递
Dynamics design of a non-natural transcription factor responding to androst-4-ene-3,17-dione 被引量:1
3
作者 Ming Zhao Mengkai Hu +6 位作者 Rumeng Han Chao Ye Xiangfei Li Tianwen Wang Yan Liu Zhenglian Xue Kun Liu 《Synthetic and Systems Biotechnology》 SCIE CSCD 2024年第3期436-444,共9页
The production of androst-4-ene-3,17-dione(AD)by the steroidal microbial cell factory requires transcription factors(TFs)to participate in metabolic regulation.However,microbial cell factory lacks effective TFs that c... The production of androst-4-ene-3,17-dione(AD)by the steroidal microbial cell factory requires transcription factors(TFs)to participate in metabolic regulation.However,microbial cell factory lacks effective TFs that can respond to AD in its metabolic pathway.Additionally,finding and obtaining natural TFs that specifically respond to AD is a complex and onerous task.In this study,we devised an artificial TF that responds to AD,termed AdT,based on structure-guided molecular dynamics(MD)simulation.According to MD analysis of the conformational changes of AdT after binding to AD,an LBD in which the N-and C-termini exhibited convergence tendencies was used as a microswitch to guide the assembly of a DNA-binding domain lexA,a linker(GGGGS)2,and a transcription activation domain B42 into an artificial TF.As a proof of design,a AD biosensor was designed and constructed in yeast on the basis of the ligand-binding domain(LBD)of hormone receptor.In addition,the transcription factor activity of AdT was increased by 1.44-fold for its variant F320Y.Overall,we created non-natural TF elements for AD microbial cell factory,and expected that the design TF strategy will be applied to running in parallel to the signaling machinery of the host cell. 展开更多
关键词 Protein design artificial transcription factor Molecular dynamics Steriods Site-mutations
原文传递
Efficient clustered,regularly interspaced short palindromic repeats-based gene activation using combinatorial human transcription activation domains
4
作者 Yi-Lian Zhou Yetong Sang +5 位作者 Lingjie Xu Chuanhong Ren Weikang Meng Yu Zhang Hongqing Liang Zehua Bao 《Protein & Cell》 2026年第2期163-169,共7页
Dear Editor,DCaLC2CO Artificial transcription factors(ATFs),which are composed of programmable DNA-binding proteins and transcription activation domains(TADs),have revolutionized the synthetic transcriptional control ... Dear Editor,DCaLC2CO Artificial transcription factors(ATFs),which are composed of programmable DNA-binding proteins and transcription activation domains(TADs),have revolutionized the synthetic transcriptional control of genes(Sang et al.,2024).The clustered,regularly interspaced short palindromic repeats(CRiSPR)systems have been repurposed as ATFs(Tanenbaum et al.,2014). 展开更多
关键词 synthetic transcriptional control genes sang CRISPR ATFs Transcription Activation Domains artificial transcription factors atfs which Gene Activation
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部