An intelligent algorithm based on neural network for combinatorial optimization problems

2014 ◽  
Author(s):  
Cui Zhang ◽  
Qiang Li ◽  
Ronglong Wang
Keyword(s):  
Neural Network ◽  
Combinatorial Optimization ◽  
Optimization Problems ◽  
Intelligent Algorithm ◽  
Combinatorial Optimization Problems

SIMULATED ANNEALING SCHEMES IN TRANSIENTLY CHAOTIC NEURAL NETWORK MODEL

2004 ◽  
Vol 18 (17n19) ◽  
pp. 2579-2584 ◽  
Author(s):  
Y. C. FENG ◽  
X. CAI
Keyword(s):  
Neural Network ◽  
Simulated Annealing ◽  
Combinatorial Optimization ◽  
Neural Network Model ◽  
Approximation Method ◽  
Optimization Problems ◽  
Chaotic Neural Network ◽  
Combinatorial Optimization Problems ◽  
Search Speed ◽  
Evolution Function

A transiently chaotic neural network (TCNN) is an approximation method for combinatorial optimization problems. The evolution function of self-back connect weight, called annealing function, influences the accurate and search speed of TCNN model. This paper analyzes two common annealing schemes. Furthermore we proposed a new subsection exponential annealing function. Finally, we compared these annealing schemes in TSP problem.

scholarly journals Deep Neural Network Approximated Dynamic Programming for Combinatorial Optimization

2020 ◽  
Vol 34 (02) ◽  
pp. 1684-1691
Author(s):  
Shenghe Xu ◽  
Shivendra S. Panwar ◽  
Murali Kodialam ◽  
T.V. Lakshman
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Dynamic Programming ◽  
Combinatorial Optimization ◽  
Optimization Problems ◽  
Computation Time ◽  
Training Procedure ◽  
Problem Size ◽  
Combinatorial Optimization Problems ◽  
Time Reduction

In this paper, we propose a general framework for combining deep neural networks (DNNs) with dynamic programming to solve combinatorial optimization problems. For problems that can be broken into smaller subproblems and solved by dynamic programming, we train a set of neural networks to replace value or policy functions at each decision step. Two variants of the neural network approximated dynamic programming (NDP) methods are proposed; in the value-based NDP method, the networks learn to estimate the value of each choice at the corresponding step, while in the policy-based NDP method the DNNs only estimate the best decision at each step. The training procedure of the NDP starts from the smallest problem size and a new DNN for the next size is trained to cooperate with previous DNNs. After all the DNNs are trained, the networks are fine-tuned together to further improve overall performance. We test NDP on the linear sum assignment problem, the traveling salesman problem and the talent scheduling problem. Experimental results show that NDP can achieve considerable computation time reduction on hard problems with reasonable performance loss. In general, NDP can be applied to reducible combinatorial optimization problems for the purpose of computation time reduction.

scholarly journals Combinatorial optimization by weight annealing in memristive hopfield networks

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Z. Fahimi ◽  
M. R. Mahmoodi ◽  
H. Nili ◽  
Valentin Polishchuk ◽  
D. B. Strukov
Keyword(s):  
Neural Network ◽  
Combinatorial Optimization ◽  
Optimization Problems ◽  
Main Idea ◽  
Hopfield Neural Network ◽  
Combinatorial Problems ◽  
Hopfield Network ◽  
Global Minima ◽  
Mixed Signal ◽  
Combinatorial Optimization Problems

AbstractThe increasing utility of specialized circuits and growing applications of optimization call for the development of efficient hardware accelerator for solving optimization problems. Hopfield neural network is a promising approach for solving combinatorial optimization problems due to the recent demonstrations of efficient mixed-signal implementation based on emerging non-volatile memory devices. Such mixed-signal accelerators also enable very efficient implementation of various annealing techniques, which are essential for finding optimal solutions. Here we propose a “weight annealing” approach, whose main idea is to ease convergence to the global minima by keeping the network close to its ground state. This is achieved by initially setting all synaptic weights to zero, thus ensuring a quick transition of the Hopfield network to its trivial global minima state and then gradually introducing weights during the annealing process. The extensive numerical simulations show that our approach leads to a better, on average, solutions for several representative combinatorial problems compared to prior Hopfield neural network solvers with chaotic or stochastic annealing. As a proof of concept, a 13-node graph partitioning problem and a 7-node maximum-weight independent set problem are solved experimentally using mixed-signal circuits based on, correspondingly, a 20 × 20 analog-grade TiO2 memristive crossbar and a 12 × 10 eFlash memory array.

Lagrangian object relaxation neural network for combinatorial optimization problems

2005 ◽  
Vol 68 ◽  
pp. 297-305 ◽  
Author(s):  
Hiroki Tamura ◽  
Zongmei Zhang ◽  
Xinshun Xu ◽  
Masahiro Ishii ◽  
Zheng Tang
Keyword(s):  
Neural Network ◽  
Combinatorial Optimization ◽  
Optimization Problems ◽  
Combinatorial Optimization Problems
Sign in / Sign up

Export Citation Format

Share Document