Learning Arbitrary Functions with Spike-Timing Dependent Plasticity Learning Rule

A neural network model based on spike-timing-dependent plasticity (STDP) learning rule, where afferent neurons excite both the target neuron and interneurons that in turn project to the target neuron, is applied to the tasks of learning AND and XOR functions. Without inhibitory plasticity, the network can learn both AND and XOR functions. Introducing inhibitory plasticity can improve the performance of learning XOR function. Maintaining a training pattern set is a method to get feedback of network performance, and would always improve network performance