The nature of statistical learning theory
The nature of statistical learning theory
The bias-variance tradeoff and the randomized GACV
Proceedings of the 1998 conference on Advances in neural information processing systems II
Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond
Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond
Diffusion Kernels on Graphs and Other Discrete Input Spaces
ICML '02 Proceedings of the Nineteenth International Conference on Machine Learning
Variable selection using svm based criteria
The Journal of Machine Learning Research
Regulatory motif finding by logic regression
Bioinformatics
Sample size for FDR-control in microarray data analysis
Bioinformatics
Classification of gene functions using support vector machine for time-course gene expression data
Computational Statistics & Data Analysis
Computational Statistics & Data Analysis
Hi-index | 0.03 |
The genetic regulatory mechanism plays a pivotal role in many biological processes ranging from development to survival. The identification of the common transcription factor binding sites (TFBSs) from a set of known co-regulated gene promoters and the identification of genes that are regulated by the transcription factor (TF) that have important roles in a particular biological function will advance our understanding of the interaction among the co-regulated genes and intricate genetic regulatory mechanism underlying this function. To identify the common TFBSs from a set of known co-regulated gene promoters and classify genes that are regulated by TFs, the new approaches using Support Vector Machine (SVM)-based Generalized Approximate Cross Validation (GACV) criteria are proposed. Two variable selection methods are considered for Recursive Feature Elimination (RFE) and Recursive Feature Addition (RFA). Performances of the proposed methods are compared with the existing SVM-based criteria, Logistic Regression Analysis (LRA), Logic Regression (LR), and Decision Tree (DT) methods by using both two real TF target genes data and the simulated data. In terms of test error rates, the proposed methods perform better than the existing methods.