From STDP towards Biologically Plausible Deep Learning

We introduce a predictive objective function for the rate aspect of spike-timing dependent plasticity (STDP), i.e., ignoring the effects of synchrony of spikes but looking at spiking rate changes. The proposed weight update is proportional to the presynaptic spiking (or firing) rate times the temporal change of the integrated postsynaptic activity. We present an intuitive explanation for the relationship between spike-timing and weight change that arises when the weight change follows this rule. Spike-based simulations agree with the proposed relationship between spike timing and the temporal change of postsynaptic activity and show a strong correlation between the biologically observed STDP behavior and the behavior obtained from simulations where the weight change follows the gradient of the predictive objective function. Finally, we draw links between this objective function, neural computation as inference, score matching, and variational EM.