Summary
International Symposium on Nonlinear Theory and Its Applications
2016
Session Number:B2L-D
Session:
Number:B2L-D-3
A Study on Performance of Hopfield-Tank Neural Networks Running on Coherent Ising Machine
Hiroyuki Yasuda, Mikio Hasegawa, Kazuyuki Aihara,
pp.-
Publication Date:2016/11/27
Online ISSN:2188-5079
Summary:
Coherent Ising Machine(CIM) which are composed with mutually interacted LASER network have been studied as one of quantum computer. CIM have been studied as ultrafast calculator for various optimization problem. It have been shown that CIM can obtain exact solution of MAX-CUT problem rapidly with high probability. We have proposed to use Hopfield-Tank Neural Network(HTNN) running on CIM to solve various optimization problems. HTNN needs some noise to obtain the exact solution of combinatorial optimization problem such as traveling salesman problem, quadratic assignment problem and so on. In this paper, we investigate performance of HTNN running on CIM with using some kinds of noises and chaotic sequences.