Biogeography-based Optimization Algorithm for Independent Component Analysis

Independent component analysis (ICA) is a signal processing technique that can be used to extract meaningful components from a dataset. Biogeography based optimization (BBO) algorithm is a recently developed stochastic optimization algorithm. In this paper, we report the use of the BBO algorithm to optimize a contrast function that measures the statistical independence of the recovered components in order to implement the ICA technique. The use of the BBO to implement the ICA technique is demonstrated on two benchmark data sets. The achieved results of using the BBO in the ICA technique are compared to that of the Fast ICA algorithm and using the particle swarm optimization (PSO) algorithm, and the differential evolutionary (DE) algorithm for ICA. Experimental results show that the BBO algorithm outperforms the Fast ICA and the DE algorithms in terms of the signal to interference ratio (SIR) of the recovered components while it outperforms the PSO algorithm in terms of the convergence speed. To both improve the convergence speed and the quality of the recovered components, the BBO and the PSO algorithms are jointly used.