Title :
Design of neural networks to tolerate the mixture of two types of faults
Author :
Tohma, Yoshihiro ; Koyanagi, Yoichi
Author_Institution :
Dept. of Comput. Sci., Tokyo Inst. of Technol., Japan
Abstract :
The authors present a design method of neural networks for optimization problems, which can tolerate the simultaneous existence of both stuck-at-zero and stuck-at-one faults. By using this new design method together with one presented earlier, neural networks can tolerate very well the mixture of the both types of faults as well as unidirectional faults.
Keywords :
neural nets; neural networks; optimization problems; stuck-at-one faults; stuck-at-zero; unidirectional faults; Application software; Computer applications; Computer networks; Computer science; Design methodology; Design optimization; Fault tolerance; Neural networks; Recurrent neural networks; Traveling salesman problems;
Conference_Titel :
Fault-Tolerant Computing, 1993. FTCS-23. Digest of Papers., The Twenty-Third International Symposium on
Conference_Location :
Toulouse, France
Print_ISBN :
0-8186-3680-7
DOI :
10.1109/FTCS.1993.627330
