DII CGO
inicioir al DIIbuscar
fcfm
top
topleft
multi-classifier systems for bioinformatics problems
 
Multi-Classifier Systems have fast been gaining popularity among researchers working in machine learning and applications for their ability to fuse together multiple classification outputs for better accuracy and classification. This talk is concerned with current issues in the design of multi-classifier systems and presents our multi-classifier developments for several Bioinformatics problems. The talk first presents some important issues in the design of multi-classifier systems, with a focus on the diversity and combination of the outputs of individual classifiers. Few diversification and combination schemes are presented. Then, a neural network based multi-classifier system for the identification of Escherichia Coli (E.Coli) promoter sequences in strings of DNA is presented. The presentation will then proceed with introducing an ensemble of neuro-fuzzy networks for micro-array cancer gene expression data classification. At the end of the talk, an analysis on the performance of several classifier fusion techniques in a protein secondary structure prediction problem will be provided. The presented approaches and results prove that ensembles of classifiers can be used as effective computational tools in solving difficult Bioinformatics problems.