Abstract
This paper presents an overlapped niulti-neural-network (OMNN). An ODINN consists of two parts: main part and partitioning part. The main part. structurally, is the same as an ordinary feedforward neural net-work, but it is considered as one consisting of several subsets. All subnets have the same input-output units. but some different hidden units. The partitioning part divides input space into several parts. each of which is associated with one subnet. An improved random search algorithm called RasID is introduced to train the OMNN. Numerical simulations show that such an OMNN has superior performance in that it has better presentation ability than an ordinary neural network and better generalization ability than a non-overlapped multi-neural-network.
