Theoretical analysis on sparsely connected associative memory networks with arbitrary degree distribution

The microcosmic topology structure of a network are known to have great influence on its macroscopic behavior. To go one step closer to biological system which displays sparsely connected architecture and complex network property, we studied a general sparely connected associative memory model using signal-to-noise analysis. Theoretical solutions for the computational performance of this model with arbitrary degree distribution are derived. Numerical simulations are carried out to demonstrate the effectiveness and rationality of the theoretical determinations, and have shown great consistence with them.