Support vector machines (SVMs) have widespread use in various classification problems. Although SVMs are often used as an off-the-shelf tool, there are still some important issues which require improvement, such as feature rescaling. Standardization is the most commonly used feature rescaling method. However, standardization does not always improve classification accuracy. This paper describes two feature rescaling methods: multiple kernel learning-based rescaling (MKL-SVM) and kernel-target alignment-based rescaling (KTA-SVM). MKL-SVM makes use of the framework of multiple kernel learning (MKL) and KTA-SVM is built upon the concept of kernel alignment, which measures the similarity between kernels. The proposed methods were compared with three other methods: an SVM method without rescaling, an SVM method with standardization, and SCADSVM. Test results demonstrate that different rescaling methods apply to different situations and that the proposed methods outperform the others in general.
microRNAs (miRNAs) have been reported to be associated with the pathogenesis and progression of breast cancer.However,little is known about the pathways through which miRNAs regulate these processes,e.g.,the interaction between miRNAs and their target genes with regard to different pathological status of breast cancer,such as histological grades.This study investigated the possible roles of miRNAs in the differentiation of histological grades of breast cancer with a computational approach.Based on a microarray dataset,15 candidate miRNAs were identified,whose predicted target genes are enriched as differentially expressed between grade I and grade III breast tumors.Among them,9 key miRNAs focalize their target genes on 6 central signaling pathways.The SMAD7 protein,the main inhibitory protein in the TGF-β pathway,is predicted as a target of several miRNAs and is also regulated by several other pathways that are possibly targeted by miRNAs.It was hypothesized that miRNAs participate in the differentiation of breast cancer and the TGF-β pathway acts as a major implementary pathway on which several miRNAs take effect through multiple channels.The prediction power of the predicted miRNA target genes was validated on three independent datasets.The differential expression of three miRNAs was validated by real-time PCR on breast carcinoma samples of 10 patients.
