To segment a image with strongly varying object sizes results generally in under-segmentation of small structures or over-segmentation of big ones, which consequences poor classification accuracies. A strategy to produce multiple segmentations of one image and classification with support vector machines (SVM) of this segmentation stack afterwards is shown.

We rederive a form of Joachims’ ξα method for tuning Support Vector Machines by the same approach as was used to derive the GACV, and show how the two methods are related. We generalize the ξα method to the nonstandard case of nonrepresentative training set and unequal misclassification costs and compare the result to the GACV estimate for the standard and nonstandard cases.

Using Support Vector Machines for MiRNA Identification

In binary classification problems, two classes of data seem to be different from each other. It is expected more complicated due the clusters in class also tend different. Traditional algorithms as Support Vector Machine (SVM) or Twin (TWSVM) cannot sufficiently exploit structural information with cluster granularity data, cause limitation on capability simulation trends. Structural (S-TWSVM) e...

In this chapter, we discuss two possible ways of improving the performance of the SVM, using geometric methods. The first adapts the kernel by magnifying the Riemannian metric in the neighborhood of the boundary, thereby increasing separation between the classes. The second method is concerned with optimal location of the separating boundary, given that the distributions of data on either side ...

X. CAO{{, J. CHEN*{, B. MATSUSHITA§, H. IMURA" and L. WANG{{ {Key Laboratory of Environment Change and Natural Disaster, Ministry of Education of China, College of Resource Science and Technology, Beijing Normal University, Beijing, 100875, China {Graduate School of Engineering, Nagoya University, Nagoya 464-8603, Japan §Graduate School of Life and Environmental Sciences, University of Tsukuba,...