神经网络实验报告.doc

《神经网络实验报告.doc》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、深圳大学实验报告实验课程名称：人工神经网络技术实验项目名称：BP神经网络对蝴蝶花分类学院：专业：软件工程报告人：学号：班级：同组人：无指导教师：实验时间：实验报告提交时间：教务处制一、实验目的初步熟悉MATLAB工作环境，熟悉命令窗口，学会使用帮助窗口查找帮助信息。二、实验内容1、网络设计，包括输入层、隐含层、输出层节点个数的设计。2、算法步骤3、编程，注意原始数据的通用化，数据输入的随机性。4、网络训练，注意训练数据与验证数据分开。5、网络验证6、结果分析，修改隐含层节点个数，修改学习率，分别对结果的影响。三、实验步骤直接在Matlab软件

2、中的Editor中运行以下代码：（完善的代码用红色字体表示）%li.m%ABPNNwithmomentumtosolveFisher'sIrisFlowerproblem%bylixiujuan,Nov13,2011%%theNNarchitecture%itisathreelayersneuralnetwork4-3-3.%%parameterdescription%h=4thenodenumerofinputlayer%i=3thenodenumerofhiddenlayer%j=3thenodenumerofoutputlayer%V[

3、h,i]theweightsbetweeninputandhiddenlayers%W[i,j]theweightsbetweenhiddenandoutputlayers%Pi[i]thethresholdsofhiddenlayernodes%Tau[j]thethresholdsofoutputlayernodes%a[h]theinputvalues%b[i]thehiddenlayernodeactivations%c[j]theoutputlayernodeactivations%ck[j]thedesiredoutputofou

4、tputlayernodes%d[j]theerorinoutputlayernodes%e[i]theerorinhiddenlayernodes%DeltaW[i,j]theamountofchangefortheweightsW[i,j]%DeltaV[h,i]theamountofchangefortheweightsV[h,i]%DeltaPi[i]theamountofchangeforthethresholdsPi[i]%DeltaTau[j]theamountofchangeforthethresholdsTau[j]%Alp

5、ha=0.1theleaningrate%Beta=0.1theleaningrate%Gamma=0.8theconstantdetermineseffectofpastweightchanges%Tor=0.001thetorrelancethatdetermineswhentostoptraining%Maxepoch=1000themaxiteratenumber%%otherparameters%Ntrain=115thenumberoftrainningsets%Ntest=35thenumberoftestsets%Otrain

6、[115]theoutputoftrainingsets%Otest[35]theoutputoftestsets%Odesired[150]thedesiredoutputoftrainingandtestsets%%functiondescription%f(x)=logsig(x)%f(x)=1/(1+exp(-x))%%datafile%inputfile:data.dat%closeall;clc;clf;clearall;%parametersfortheNNstructureh=4;i=3;j=3;Alpha=0.1;Beta=

7、0.1;Gamma=0.85;Tor=0.0005;Maxepoch=2000;Accuracy=0;Ntrain=115;Ntest=35;%assignrandomvaluesintherange[-1,+1]V=2*(rand(h,i)-0.5);W=2*(rand(i,j)-0.5);Pi=2*(rand(1,i)-0.5);Tau=2*(rand(1,j)-0.5);DeltaWOld(i,j)=0;%setthedelatofWijto0DeltaVOld(h,i)=0;%setthedelatofVijto0DeltaPiOld

8、(i)=0;%setthedelatofPito0DeltaTauOld(j)=0;%setthedelatofTauto0%thelearningprocessE