System neural network learning.

摘要

DESCRIBED LEARNING SYSTEM NEURAL NETWORK IN WHICH HAS BEEN AN INTERFERED input-output relationship. THE LEARNING SYSTEM NETWORK NEURAL includes a portion (12) PROBABILITY DENSITY FOR DETERMINING A PROBABILITY DENSITY ON SPACE SUM OF SPACE ENTRY AND SPACE OUT FROM A SET OF SAMPLES IN AND OUT GIVEN BY LEARNING It is defined probability density on Space SUMA to have a parameter, and a portion (13) INTERFERENCE FOR DETERMINING probability density function based upon the probability density FROM PART OF pROBABILITY DENSITY SO THAT THE RELATIONSHIP IN-OUT OF sAMPLES IS INTERREFERIDA A FINDS FROM probability density function HAVING A dETERMINATION parameter value by LEARNING TO BE REPEATED LEARNING setting until VALUE OF DIFFERENTIAL fUNCTION pARAMETER DEFAULT PROBABILITY using a method previously described MAXIMA IS SMALLER THAN A VALUE OF RE PREVIOUSLY DESCRIBED ference.