library(AMORE)
data<-read.table('G:\\dataguru\\ML\\ML09\\基于BP网络的个人信贷信用评估\\基于BP网络的个人信贷信用评估\\german.data-numeric')
for (i in 1:25) {
data[,i] <- as.numeric(as.vector(data)[,i])
}
pos<-data[which(data$V25=='1'),]
neg<-data[which(data$V25=='2'),]
train<-rbind(pos[1:350,],neg[1:150,])
test<-rbind(pos[351:700,],neg[151:300,])
net <- newff(n.neurons=c(24,8,2,1), learning.rate.global=1e-13, momentum.global=0.5,
error.criterium="LMS", Stao=NA, hidden.layer="tansig",
output.layer="purelin", method="ADAPTgdwm")
result <- train(net, train[1:24], train[25], error.criterium="LMS", report=TRUE, show.step=100, n.shows=5 )
y <- sim(result$net, test[1:24])
y[which(y<1.5)] <- 1
y[which(y>=1.5)] <- 2
sum = 0
for(i in 1:500){
if(y[i]==test[i,25]){
sum =sum+1
}
}
cat("正确率", sum/500, "n")