Support Vector Machines (SVMs) have been very successful in text classification. However, the intrinsic geometric structure of text data has been ignored by standard kernels commo...
As tools and systems for producing and disseminating image data have improved significantly in recent years, the volume of digital images has grown rapidly. An efficient mechanism ...
Abstract. Through task decomposition and module combination, minmax modular support vector machines (M3 -SVMs) can be successfully used for difficult pattern classification task. ...
Abstract. Single classifiers, such as Neural Networks, Support Vector Machines, Decision Trees and other, can be used to perform classification of data for relatively simple proble...
George L. Tsirogiannis, Dimitrios S. Frossyniotis,...
We study the problem of classifying images into a given, pre-determined taxonomy. The task can be elegantly translated into the structured learning framework. Structured learning, ...