Correlation matrices and the construction of successful network recodings

Previous work performed by the authors demonstrated the importance of the amount of synaptic connectivity and output firing thresholds, and the relative unimportance of an activity based weight modification rule, in the construction of neural networks which perform successful recodings. The authors attempt to construct near optimal recoding networks incorporating and building upon the results of previous work. In one series of simulations, they generalize the previous results on optimal network connectivity levels. In a second series of simulations, they illustrate the desirability of a mechanism that allows individual setting of the afferent connection weights and postsynaptic neuron firing threshold. Additionally, the success of an approximation technique in the determination of optimal synaptic weights and firing thresholds is demonstrated