LOCAL CLUSTER NEURAL-NET ANALOG VLSI DESIGN

The local cluster (LC) artificial neural net is a special kind of mult ilayer perceptron where the sigmoid functions combine in clusters that have a localised response in input space. The proponents of the LC ar chitecture have shown that it is versatile and trains well. They also suggested that the LC nets could be suitable for realisation in analog VLSI. We investigated the feasibility of an analog realisation of LC nets by following through the complete cycle from design to fabricatio n. We found that all the required mathematical functions for an analog realisation of LC net can be realised in current mode bipolar and CMO S circuits. In this paper we discuss the main design issues paying spe cial attention to the alternative training regimes for a LC chip. (C) 1998 Elsevier Science B.V. All rights reserved.