ICA model estimation using a mixed learning rule based on neural network and genetic algorithm
Doru Constantin
University of Piteşti, Piteşti, Romania
Abstract:
This paper presents a new method of estimating the independent component analysis model based on the use of a training algorithm based on a mixed learning algorithm of neural networks type with genetic algorithms. The mixed training algorithm is applied to optimize the objective function used to estimate the ICA model. The experimental results for the new independent component estimation algorithm are established in specific blind source separation applications. The presented results are a joint work with Costel Bălcău.