Description: 用GA直接训练BP网络的权重算法
主程序:gafault.m 它包括以下子程序:
1. BP网络初始化:nninit.m――给出P,T,R,S1,S2;
2. 适应值计算函数:gabpEval.m;
3.将遗传算法的编码解码为BP网络所对应的权值、阈值函数:gadecod.m;
-GA directly with BP network training algorithm the weights of the main program: gafault.m It includes the following subroutines: 1. BP network initialization: nninit.m-- given P, T, R, S1, S2 2. Adaptation value function: gabpEval.m 3. Genetic algorithm for BP network codec corresponding to the right value, the threshold function: gadecod.m Platform: |
Size: 1024 |
Author:QAP |
Hits:
Description: bp算法实现代码,在c++ builder 平台上实现. b p算法实现代码,在c++ builder 平台上实现.
-bp algorithm code, in c++ builder platform. bp algorithm code, in c++ builder platform. Platform: |
Size: 460800 |
Author:杨林 |
Hits:
Description: 一、用GA直接训练BP网络的权重算法
主程序:gafault.m 它包括以下子程序:
1. BP网络初始化:nninit.m――给出P,T,R,S1,S2;
2. 适应值计算函数:gabpEval.m;
3.将遗传算法的编码解码为BP网络所对应的权值、阈值函数:gadecod.m;
二、用GA先求BP网络的权重,再用纯BP直接训练BP的混合GA-BP算法
主程序:gabpfault.m 它包括以下子程序:
1. 网络初始化:nninit.m――给出P,T,R,S1,S2;
2. 适应值计算函数:gabpEval.m;
3.将遗传算法的编码解码为BP网络所对应的权值、阈值函数:gadecod.m;
三、纯BP
主程序:(1)bpfault.m 在MATLAB5.2上
(2)bpfault.m 在MATLAB6.5上 为后来所加
-First, the direct training of BP network with GA the weight algorithm
Main program: gafault.m It includes the following routines:
1. BP network initialization: nninit.m-- given P, T, R, S1, S2
2. Fitness calculation functions: gabpEval.m
3. Of genetic algorithms for the BP network codec corresponding weights, the threshold function: gadecod.m
Second, with the GA first aim at the weight of BP network, and then direct the training of pure BP mixture of BP algorithm GA-BP
Main program: gabpfault.m It includes the following routines:
1. Network initialization: nninit.m-- given P, T, R, S1, S2
2. Fitness calculation functions: gabpEval.m
3. Of genetic algorithms for the BP network codec corresponding weights, the threshold function: gadecod.m
3, pure BP
Main program: (1) bpfault.m on the MATLAB5.2
(2) bpfault.m the MATLAB6.5 for the subsequently added Platform: |
Size: 38912 |
Author:zhanghr |
Hits:
Description: 利用BP网络实现函数逼近,本程序以cos(k*pi*p)为例进行逼近-Function approximation using BP network, the procedures to cos (k* pi* p) as an example approximation Platform: |
Size: 1024 |
Author:tanjing |
Hits:
Description: 用动量梯度下降法训练BP网络
已知输入向量为P=[-1,-2,3,1 -1,1,5,3],目标输出为T=[-1,-1,1,1]。
-Gradient descent with momentum BP network training input vector is known as P = [-1,-2,3,1 -1,1,5,3], the target output for the T = [-1,-1,1, 1]. Platform: |
Size: 59392 |
Author:张玲 |
Hits:
Description: C++和C编写的bp神经网络源程序,对bp神经网络的学习很有用-C++ and C source code written in bp neural network,it is useful to bp neural network learning Platform: |
Size: 12288 |
Author:洛龙行 |
Hits:
Description: GA优化BP神经网络初始权值,阈值,从而增强其鲁棒性(GA optimizes the initial weights and thresholds of BP neural network, thereby enhancing its robustness.) Platform: |
Size: 186368 |
Author:Wyy_999 |
Hits: