A simple, homogeneous parallel PCA network

A form of ANN using interneurons has been shown to be capable of performing a principal component analysis of the input data. A review is given of the algorithm which does this. A model of the interneuron network with a set of more biologically feasible initial conditions has been developed-the weights to and from each interneuron are independent and yet the weights to and from the interneurons are shown to perform a PCA. A new parallel algorithm is proposed using the innate properties of the network. This is shown to converge in a completely parallel and homogeneous fashion to the principal components.