Time Efficient Optimal Tunning through Various Learning Rules in Unify Computing (TEOTLRUC)

In this paper the effects of learning rules on mutual synchronization of various tree parity machines are presented. The experiment of tree parity machine is not confined to single hidden layer machine. Machines with double and triple hidden layers are also have been studied. Basically a tree parity machine has a single hidden layer. Once two tree parity machine of same size are created then they both are synchronized mutually in order to obtain same weight vectors. If outputs are identical of both the machines then a suitable learning rule is applied to the neural weights of both tree parity machines, in order to generate identical weight values. In this paper the mutual synchronization is examined on all three tree parity machines with different number of hidden layers. Results are compared in terms mutual synchronization time, which shows results for the proposed system.