Solving Optimization Problems Based on Chaotic Neural Network with Hysteretic Activation Function

2005 ◽  
pp. 745-749
Author(s):  
Xiuhong Wang ◽  
Qingli Qiao ◽  
Zhengqu Wang
Keyword(s):  
Neural Network ◽  
Optimization Problems ◽  
Activation Function ◽  
Chaotic Neural Network

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 PRNG Implementation Based on Chaotic Neural Network (CNN)

2019 ◽  
Vol 8 (4) ◽  
pp. 158
Author(s):  
Mohammed J. Mohammed
Keyword(s):  
Neural Network ◽  
Random Number ◽  
Random Number Generator ◽  
Activation Function ◽  
Binary Number ◽  
Chaotic Neural Network ◽  
Adaptive Architecture ◽  
Large Numbers ◽  
Pseudo Random Number Generator ◽  
Randomness Tests

In this work, a neural network with chaos activation function has been applied as a pseudo-random number generator (PRNG). Chaotic neural network (CNN) is used because of its noise like behaviour which is important for cryptanalyst to know about the hidden information as it is hard to predict. A suitable adaptive architecture was adopted to generate a binary number and the result was tested for randomness using National Institute of Standard Technology (NIST) randomness tests. Although the applications of CNN in cryptography have less effective than traditional implementations, this is because these systems need large numbers of digital logic or even a computer system. This work will focus on applications that can use the proposed system in an efficient way that minimize the system complexity.

Noise Compensation Based Hysteretic Noisy Chaotic Neural Network for Broadcast Scheduling Problem in Wireless Multihop Networks

2012 ◽  
Vol 151 ◽  
pp. 527-531
Author(s):  
Ming Sun ◽  
Yuan Guo ◽  
Xue Feng Dai ◽  
Yao Qun Xu
Keyword(s):  
Neural Network ◽  
Activation Function ◽  
Wireless Multihop Networks ◽  
Scheduling Problem ◽  
Broadcast Scheduling ◽  
Noise Levels ◽  
Chaotic Neural Network ◽  
Noise Compensation ◽  
Compensation Factor ◽  
Lower Noise

Compared with the noisy chaotic neural network, hysteretic noisy chaotic neural network always exhibits better optimization performance at higher noise levels, but exhibits worse optimization performance at lower noise levels. In order to enable the hysteretic noisy chaotic neural network to behave more excellent optimization performance not only at higher noise levels but also at lower noise levels, we introduce a noise compensation factor to the original hysteretic noisy chaotic neural network, and present noise compensation based hysteretic noisy chaotic neural network. The proposed network can outperform the hysteretic noisy chaotic neural network by the interaction of hysteretic activation function and the noise compensation factor. One benchmark broadcast scheduling problem is used to verify the superiority of the proposed network. The simulation results show that the proposed network takes advantages over the noisy chaotic neural network, the hysteretic noisy chaotic neural network and other algorithms.

A Novel Chaotic Neural Network with Radial Basis Function and their Application to TSP

2012 ◽  
Vol 151 ◽  
pp. 532-536
Author(s):  
Nan Xu ◽  
Li Jie Liu ◽  
Yao Qun Xu
Keyword(s):  
Neural Network ◽  
Radial Basis Function ◽  
Basis Function ◽  
High Speed ◽  
Optimization Problem ◽  
Activation Function ◽  
Chaotic Neural Network ◽  
Combinational Optimization ◽  
Radial Basis ◽  
Sigmoid Activation Function

A novel transiently chaotic neural network with radial basis function is proposed by analyzing the capability of chaotic search and the effect in solving combinational optimization problem. The character of Radial basis function is higher nonlinear and better function approaching ability. So a novel transiently chaotic neural network is presented by transferring sigmoid activation function into non-monotonous activation function which is composed by Contrary Multiquadric function and Sigmoid. This network is used to solve Traveling Salesman Problem (TSP), and the simulation result indicates that it can avoid the limit of being trapped into the local minima and converge to the global minima with high speed.

Sign in / Sign up

Export Citation Format

Share Document