Learning Using Soft Computing Techniques
V. Susheela Devi · Advances in computational intelligence and robotics book series · 2013
This chapter focuses on a few key applications in the field of classification and clustering. Techniques of soft computing have been used to solve these applications. The first application finds a new similarity measure for time series data, combining some available similarity measures. The weight to be given to these similarity measures is found using a genetic algorithm. The other applications discussed are for pattern clustering. A Particle Swarm Optimization (PSO) algorithm has been used for clustering. A modification of the PSO using genetic operators has been suggested. In addition, simultaneous clustering and feature selection and simultaneous clustering and feature weighting has been discussed. Results have been given for all the techniques showing the improvement achieved using these techniques.