Parallel, self-organizing, hierarchical neutral networks with continuous inputs and outputs

Parallel, self-organizing, hierarchical neural networks (PSHNNs) are multistage networks in which stages operate in parallel rather than in series during testing. Previous PSHNNs assume quantized, say, binary outputs. A new type of PSHNN is discussed such that the outputs are allowed to be continuous-valued. The performance of the resulting network is tested in the problem of predicting speech signal samples from past samples. Both delta rule and sequential least squares learning are used. In all cases studied, the new network achieves better performance than linear prediction.<>