supervised associative learning in spiking neural network脉冲神经网络中的有监督联想学习.pdfVIP

Supervised Associative Learning in Spiking Neural Network Nooraini Yusoff and André Grüning, Department of Computing, Faculty of Engineering and Physical Sciences, University of Surrey, Guildford, GU2 7XH Surrey, UK {n.yusoff, a.gruning}@surrey.ac.uk Abstract. In this paper, we propose a simple supervised associative learning approach for spiking neural networks. In an excitatory-inhibitory network paradigm with Izhikevich spiking neurons, synaptic plasticity is implemented on excitatory to excitatory synapses dependent on both spike emission rates and spike timings. As results of learning, the network is able to associate not just familiar stimuli but also novel stimuli observed through synchronised activity within the same subpopulation and between two associated subpopulations. Keywords: Spiking neural network, Associative learning, Supervised learning, Excitatory-Inhibitory network, Izhikevich spiking neurons. 1 Introduction There is evidence in neurophysiology that long-term association between stimuli, which involves synaptic plasticity, is triggered by overlapping short-term activity, which only involves activity dynamics (e.g. [1], [7] and [8]), linking neuronal activity and long-term memory. Associative-based learning can be implemented using unsupervised or supervised approaches [6]. For unsupervised learning, perhaps the temporal variant of Hebbian learning known as spike-timing dependent plasticity (STDP) is the most biologically plausible approach. However, the purely unsupervised approach is often not suitable for goal-oriented applications, so it is used in this paper with prescribed target s