Study on DBM network with non-monotonic neurons

In this paper, we report a study on learning ability of a Deterministic Boltzmann Machine (DBM) with neurons which have a non-monotonic activation function. We use an end-cut-off-type function with a threshold parameter `θ' as the non-monotonic function. Numerical simulations of learning nonlinear problems, such as the XOR problem and the ADD problem, show that the DBM network with non-monotonic neurons has higher learning ability compared to the network with monotonic neurons, and that the non-monotonic neural network has a novel effects which adjust the number of neurons. We have designed an integrated circuit of the 2-3-1 DBM network. The use of the non-monotonic neurons make it possible to integrate a large scale neural network because of the simple circuit design.