A Reinforcement Learning Framework for Spiking Networks with Dynamic Synapses
Figure 3
Simulation results in case of five neurons in the hidden layer and window size set to five msec. (a) Values of reward signal. (b) Distances between the reference and the output signal, . (c) Maximum cross-correlation coefficient observed between the reference and the output signal, . A snapshot from the simulation over the input/output firing patterns and internal EPSP of the output neuron is given in Figure S1.
