Location:
Search - bp x
Search list
Description: 本文件为用C语言实现的可实现广义异或问题的bp神经网络算法。该问题是对标准异或问题的推广。在标准异或问题中，输入X1和X2取离散量-1或+1，在广义异或问题中，输入（X1,X2）可以在区间[-1，+1] X [-1, +1]内任意取值，而输出为Y=sign(x1,x2)，其中sign()为符号函数，在区间[-1，+1] X [-1, +1]内随机产生500个训练样本.本程序用标准BP网实现该分类问题.-this document for the use of C language of achieving broad differences or problems bp neural network algorithm. The problem is different to the standard or the promotion. Differences in standards or issue, the importation of X1 and X2 from discrete or a -1, in the broadest sense of differences or problems, the input (X1, X2) in the interval [-1, 1] X [1, 1] arbitrary value. while output Y = sign (x1, x2), in which the sign () to sign function in the interval [-1, 1] X [1, 1] with randomly generated 500 samples. the standard procedure used in the BP Network classification.
Platform: |
Size: 15954 |
Author: 刘波 |
Hits:
Description: BP算法拟合f(x)=sin(x)函数,样本数为9对，目标值0.001-BP fitting algorithm f (x) = sin (x) function, sample size for the nine right, the target value 0.001
Platform: |
Size: 14900 |
Author: 孙志海 |
Hits:
Description: BP算法拟合f(x1,x2)=sin((x1)/x1)*sin((x2)/x2)*函数,样本数为21*21对-BP fitting algorithm f (x1, x2) = sin ((x) / x) * sin ((x2) / x2) * function, sample size of 21 * 21 pairs
Platform: |
Size: 17293 |
Author: 孙志海 |
Hits:
Description: zemax源码：
This DLL models an anamorphic aspheric surface.
This surface is essentially an even aspheric surface with different terms for
the x and y directions.
The sag is given by:
Z = ((CX*x*x)+(CY*y*y)) / (1 + sqrt(1-((1+KX)*CX*CX*x*x)-((1+KY)*CY*CY*y*y)))
+ AR*( (1 - AP)*x*x + (1 + AP)*y*y )^2
+ BR*( (1 - BP)*x*x + (1 + BP)*y*y )^3
+ CR*( (1 - CP)*x*x + (1 + CP)*y*y )^4
+ DR*( (1 - DP)*x*x + (1 + DP)*y*y )^5
Note the terms AR, BR, CR, and DR ... have units of length to the -3, -5, -7, and -9 power.
The terms AP, BP, CP, and DP are dimensionless.
The surface is rotationally symmetric only if AP = BP = CP = DP == 0 and CX = CY and KX = KY.
Platform: |
Size: 23995 |
Author: 狄拉克 |
Hits:
Description: BP算法
给定一个由N维向量X组成的集合，这些向量将是一个响应agent的感知处理单元计算出的特征向量。这些分量的值可以是数值，也可以是布尔值。这些动作也许是学习者所观察到的一个教师对一组输入的响应。这些相关的动作有时称为向量的“标号”或“类别”。集合与响应的标号组成“训练集合(training set)”.机器学习问题就是寻找一个函数。
Platform: |
Size: 1227 |
Author: 张志丹 |
Hits:
Description: 用bp神经网络对sin(x)拟合,隐含层采用sigmoid函数，输出层采用线性函数,
Platform: |
Size: 1285 |
Author: jesse |
Hits:
Description: 本文件为用C语言实现的可实现广义异或问题的bp神经网络算法。该问题是对标准异或问题的推广。在标准异或问题中，输入X1和X2取离散量-1或+1，在广义异或问题中，输入（X1,X2）可以在区间[-1，+1] X [-1, +1]内任意取值，而输出为Y=sign(x1,x2)，其中sign()为符号函数，在区间[-1，+1] X [-1, +1]内随机产生500个训练样本.本程序用标准BP网实现该分类问题.-this document for the use of C language of achieving broad differences or problems bp neural network algorithm. The problem is different to the standard or the promotion. Differences in standards or issue, the importation of X1 and X2 from discrete or a-1, in the broadest sense of differences or problems, the input (X1, X2) in the interval [-1, 1] X [1, 1] arbitrary value. while output Y = sign (x1, x2), in which the sign () to sign function in the interval [-1, 1] X [1, 1] with randomly generated 500 samples. the standard procedure used in the BP Network classification.
Platform: |
Size: 52224 |
Author: 刘波 |
Hits:
Description: 一个用delphi写的基于bp网络逼近函数的方法的源程序，程序提供的逼近函数包括：sin(x)*exp(-x)、sin(x)/x\1/(1+exp(-x))、(1-exp(-x))/(1+exp(-x))、阶跃、高斯。~..~
-with a written bp- based networks function approximation of the original source code. procedures for the approximation functions include : sin (x)* exp (-x), sin (x)/x \ 1/(1 exp (-x)), (1- exp (-x))/(1 exp (-x)), Step, Gaussian. ~ ~ ..
Platform: |
Size: 308224 |
Author: q024100404 |
Hits:
Description: BP算法拟合f(x)=sin(x)函数,样本数为9对，目标值0.001-BP fitting algorithm f (x) = sin (x) function, sample size for the nine right, the target value 0.001
Platform: |
Size: 61440 |
Author: 孙志海 |
Hits:
Description: BP算法拟合f(x1,x2)=sin((x1)/x1)*sin((x2)/x2)*函数,样本数为21*21对-BP fitting algorithm f (x1, x2) = sin ((x)/x)* sin ((x2)/x2)* function, sample size of 21* 21 pairs
Platform: |
Size: 17408 |
Author: 孙志海 |
Hits:
Description: BP算法拟合y=abs(sin(x))函数，样本数9对，目标值0.001-BP algorithm fitting y = abs (sin (x)) function, for nine samples right, the target value 0.001
Platform: |
Size: 61440 |
Author: 孙志海 |
Hits:
Description: zemax源码：
This DLL models an anamorphic aspheric surface.
This surface is essentially an even aspheric surface with different terms for
the x and y directions.
The sag is given by:
Z = ((CX*x*x)+(CY*y*y)) / (1 + sqrt(1-((1+KX)*CX*CX*x*x)-((1+KY)*CY*CY*y*y)))
+ AR*( (1 - AP)*x*x + (1 + AP)*y*y )^2
+ BR*( (1 - BP)*x*x + (1 + BP)*y*y )^3
+ CR*( (1 - CP)*x*x + (1 + CP)*y*y )^4
+ DR*( (1 - DP)*x*x + (1 + DP)*y*y )^5
Note the terms AR, BR, CR, and DR ... have units of length to the -3, -5, -7, and -9 power.
The terms AP, BP, CP, and DP are dimensionless.
The surface is rotationally symmetric only if AP = BP = CP = DP == 0 and CX = CY and KX = KY.
-ZEMAX source: This DLL models an anamorphic aspheric surface.This surface is essentially an even aspheric surface with different terms forthe x and y directions.The sag is given by: Z = ((CX* x* x)+ (CY* y* y))/(1+ sqrt (1- ((1+ KX)* CX* CX* x* x)- ((1+ KY)* CY* CY* y* y)))+ AR* ((1- AP)* x* x+ (1+ AP)* y* y) ^ 2+ BR* ((1- BP)* x* x+ (1+ BP)* y* y) ^ 3+ CR* ((1- CP)* x* x+ (1+ CP)* y* y) ^ 4+ DR* ((1- DP)* x* x+ (1 2B ! DP)* y* y) ^ 5Note the terms AR, BR, CR, and DR ... have units of length to the-3,-5,-7, and-9 power.The terms AP, BP, CP , and DP are dimensionless.The surface is rotationally symmetric only if AP = BP = CP = DP == 0 and CX = CY and KX = KY.
Platform: |
Size: 23552 |
Author: 狄拉克 |
Hits:
Description: 这是一个用BP神经网络解决XOR问题的VC程序，楞可以通过修改初始化参数，可将该程序用于其它分类的问题-this is a BP neural network to solve the XOR problem VC procedures, Labrang may amend initialization parameters, this program can be used for other classifications of
Platform: |
Size: 51200 |
Author: hongbochen |
Hits:
Description: 用vc＋＋实现的BP神经网络解决异或问题的源代码。非常经典。-Using vc realize the BP neural network to solve XOR problem
Platform: |
Size: 88064 |
Author: wxr |
Hits:
Description: 用BP算法拟合三角函数y=0.4sin(2πx)+0.5-BP algorithm fitting with trigonometric functions y = 0.4sin (2πx)+ 0.5
Platform: |
Size: 978944 |
Author: 罗肖 |
Hits:
Description: matlab平台上实现函数y=sinx的逼近程序，构造bp算法实现（输出层采用y=x线性函数，隐含层采用非对称Sigmoid函数）-matlab platform function y = sinx approximation procedures, construction bp algorithm (output layer using linear function y = x, hidden layer use of non-symmetrical Sigmoid function)
Platform: |
Size: 1024 |
Author: 黄泉 |
Hits:
Description: BP算法
给定一个由N维向量X组成的集合，这些向量将是一个响应agent的感知处理单元计算出的特征向量。这些分量的值可以是数值，也可以是布尔值。这些动作也许是学习者所观察到的一个教师对一组输入的响应。这些相关的动作有时称为向量的“标号”或“类别”。集合与响应的标号组成“训练集合(training set)”.机器学习问题就是寻找一个函数。-BP algorithm given by the N-dimensional vector X composed of a collection, these vectors will be a response to the perceived agent processing unit eigenvector calculated. These components can be numerical values may also be a boolean value. These movements may be observed by the learner, a teacher of a group of input response. These related actions are sometimes referred to as Vector s labeling or category. Collection with the label in response to form a training set (training set) . Machine learning problem is to find a function.
Platform: |
Size: 1024 |
Author: 张志丹 |
Hits:
Description: 用bp神经网络对sin(x)拟合,隐含层采用sigmoid函数，输出层采用线性函数,-Bp neural network used to sin (x) fitted to the use of hidden layer sigmoid function, using a linear function of output layer,
Platform: |
Size: 1024 |
Author: jesse |
Hits:
Description: BP算法解决函数y=cos( 2*PI* x )学习问题-BP algorithm to solve function y = cos (2* PI* x) learning problems
Platform: |
Size: 26624 |
Author: 孙胖 |
Hits:
Description: 第一步：运行BP_CCPP.m
得到并保存传统BP预测，初始权值和阈值w1_BP(w1 B1 w2 B2)，BP网络(BP_net)，和结果。
第二部：运行GABP_CCPP.m
得到并保存GABP预测，初始权值和阈值(x)，优化后的权值和阈值w1_GABP(w1 B1 w2 B2)，BP网络(GA_BP_net)，和结果。
第三部：运行sum_BP.m
用第一步和第二部得到的网络分别预测，得到结果对比和图形。
第四部：运行WB.m
得到部分权值和阈值图。(First step: run BP_CCPP.m Get and save traditional BP prediction, initial weight and threshold value w1_BP (W1 B1 W2 B2), BP network (BP_net), and results. The second part: running GABP_CCPP.m Get and save GABP prediction, initial weight and threshold (x), optimized weight and threshold value and threshold w1_GABP (W1 B1 W2 B2), BP network (GA_BP_net), and results. The third part: running sum_BP.m The first and second networks are used to predict the results, and the results are compared and the figures are obtained. The fourth part: running WB.m Get some weight and threshold graph.)
Platform: |
Size: 8531968 |
Author: yangyanghi0 |
Hits: