9.1、神经网络实例

神经网络实例

junjun

2016年2月10日

实例一、神经网络用于鸢尾花

nnet包的nnet函数，nnet只能创建single-hidden-layer的神经网络

#1、加载数据data("iris")#2、创建训练集和测试集数据index <- sample(1:2, nrow(iris), prob=c(0.7, 0.3), replace = T)train_iris <- iris[index==1, ]test_iris <- iris[index==2, ]#3、建模library(nnet)model <- nnet(Species~., train_iris, size=20, decay=0.01, maxit=500, trace=F)#4、模型评估model

## a 4-20-3 network with 163 weights## inputs: Sepal.Length Sepal.Width Petal.Length Petal.Width ## output(s): Species ## options were - softmax modelling  decay=0.01

pred <- predict(model, train_iris, type="class")mean(pred==train_iris[, 5])

## [1] 0.9807692

table(pred, train_iris[, 5])

##             ## pred         setosa versicolor virginica##   setosa         39          0         0##   versicolor      0         25         1##   virginica       0          1        38

#5、预测pred_iris <- predict(model, test_iris, type="class")mean(pred_iris==test_iris[, 5])

## [1] 1

table(pred_iris, test_iris[, 5])

##             ## pred_iris    setosa versicolor virginica##   setosa         11          0         0##   versicolor      0         24         0##   virginica       0          0        11