In this work, we show that by representing the neural activity and synaptic activity in the complex domain, it is possible to encode temporal relations among the neural oscillations. The complex representation is also able to bring out learning property like Spike-Timing Dependent Plasticity (STDP) observed in a network of neurons.