【摘 要】
：
Defining informative features from complex and high dimensional biological data is of great importance in disease study,drug development,etc.Support vector machine-recursive feature elimination(SVM-RF
【出 处】
：
第二届中国计算机学会生物信息学会议
论文部分内容阅读
Defining informative features from complex and high dimensional biological data is of great importance in disease study,drug development,etc.Support vector machine-recursive feature elimination(SVM-RFE)is a very popular data analysis technique and has shown its power in many fields.It ranks the features according to the recursive deletion sequence based on SVM.This paper studies the criteria to determine how many top ranked features should be selected according to the recursive feature elimination procedure of SVM-RFE.The classification accuracy rate and overlapping ratio of the samples on the current subspace are adopted to measure the discriminative ability of the feature subset.
其他文献
Most proteins perform their biological functions while interacting as complexes.The detection of protein complexes is an important task not only for understanding the relationship between functions an
Intelligent optimization algorithms have advantages in dealing with complex nonlinear problems accompanied by good flexibility and adaptability.In this paper,the FCBF(Fast Correlation-Based Feature se
Analysis of large-scale gene expression data is a research hotspot in the field of bioinformatics,which can be used to diagnose the disease of human and animal,and to study the abnormal phenomenon in
利用双聚类算法在大规模基因表达数据上进行聚类分析可以发现不同的癌症亚型,结合基因网络数据可以提高癌症亚型分类的准确度。已有整合网络的双聚类算法通常仅基于基因的度加权选择基因,易受网络中噪声互作的干扰和缺失互作的误导。为此,本文提出了一种基于基因网络正则化的双聚类算法(Network Regularized Bi-Clustering algorithm,NetRBC)。
In this paper,a Hepatitis B virus(HBV)model with an incubation period,and delayed state and control variables is firstly proposed; furthermore the combination treatment is adopted in order to have a l
The prediction of residue solvent accessibility(RSA)can provide more information for analyzing protein structures and functions.Many computing methods have been proposed to predict it for better perfo
Simulating multi-scale dynamics of complex living systems is the major challenge in the researches of computational system biology.In this work,we propose a CUDA-based generic multi-cellular biologica
Ribosome stalling is manifested by the local accumulation of ribosomes at specific codon positions of mRNAs.Here,we present ROSE,a deep learning framework to analyze high-throughput ribosome profiling
Docker 应用容器引擎可实现打包生物信息数据流应用程序以及依赖包到一个可移植的容器中,然后部署到任何主流的 Linux 机器上。本实验室利用Docker 技术结合make 搭建面向RNA-Seq、全基因组重测序、Pacbio 三代全长转录组测序等生物信息分析软件工作流程的Docker 容器。产出的大型工作流可以实现RNA-Seq 表达差异分析及GO、KEGG 等相关注释分析,同时能实现对fus
Multi-view classification and feature selection have received considerable attention in recent years.In many real classification problems,the data in each view may have noise.The low-rank regression m