Learning probabilistic RAM nets using VLSI structures

Hardware-realizable learning probabilistic RAMs (pRAMs) which implement local reinforcement rules utilizing synaptic rather than threshold noise in the stochastic search procedure are described. The design allows for both global and local rewards and penalties (in this latter case implementing a modified version of backpropagation). The architecture allows for serial updating of the weights of a pRAM net according to a reward/penalty learning rule. It is possible to generate a new set of pRAM outputs at least every 100 μs, which is faster than the response time of biological neurons