CISUC

Biological Data Analysis using Incremental Kernel Machines: A Comparative Study

Authors

Abstract

Kernel Machines, and Support Vector Machines in particular, are state-of-the-art performers in the current machine learning paradigm. However, traditional implementations work in a batch fashion, limiting its application. Recent developments conducted to incremental algorithms that further decompose the learning process, which enable the application to large scale datasets, a common feature in biological data analysis. This work presents a set of experiments in order to evaluate the performance of incremental algorithms using five biological benchmark datasets, opening the possibility to further extend these incremental approaches to large scale problems.

Subject

kernels machines;Biomedical data mining

Conference

WACIâ??08 - Workshop on Applications of Computational Intelligence, December 2008


Cited by

No citations found