×

Evolutionary conditional rules versus support vector machines weighted formulas for liver fibrosis degree prediction. (English) Zbl 1212.68152

Summary: Present paper brings together two novel evolutionary techniques designed for classification and applied for the differentiation among five possible degrees of liver fibrosis within chronic hepatitis C. A purely evolutionary method – the cooperative coevolutionary classifier – endowed with a hill climbing algorithm for the selection of influential attributes is put in opposition to a hybridized approach for the task – the evolutionary support vector machine. Each of the two exhibit interesting resulting features as regards additional information on the importance of each indicator and the interaction among these for the final predicted out come. The medical experts can eventually benefit from both methodologies as a support for their decision making and decide what further knowledge they need to extract from them, i.e., either in the form of conditional rules, weighted formulas or both.

MSC:

68T05 Learning and adaptive systems in artificial intelligence
68T20 Problem solving in the context of artificial intelligence (heuristics, search strategies, etc.)
PDFBibTeX XMLCite