In machine learning problems with tens of thousands of features and only dozens or hundreds of independent training examples, dimensionality reduction is essential for good learni...
Speech recognition is usually based on Hidden Markov Models (HMMs), which represent the temporal dynamics of speech very efficiently, and Gaussian mixture models, which do non-opt...
Distance-based methods in machine learning and pattern recognition have to rely on a metric distance between points in the input space. Instead of specifying a metric a priori, we...
Background: The current progress in sequencing projects calls for rapid, reliable and accurate function assignments of gene products. A variety of methods has been designed to ann...
Abstract: Kernel classifiers based on Support Vector Machines (SVM) have achieved state-ofthe-art results in several visual classification tasks, however, recent publications and d...
Guo ShengYang, Min Tan, Si-Yao Fu, Zeng-Guang Hou,...