如何解决我面临的一个问题是具有多类的SVM分类错误说我最小没有意义
我正在为12个元音绘制一个SVM模型。将有12节课。我的文件是https://docs.google.com/spreadsheets/d/1eXQ7mq7bm8yFdBaUrijCUsLGS9nFlhb37esA8p61ZyM/edit?usp=sharing
我的代码是
#Preparing the variables
dfRead <- read.csv("VTLN04.csv")
dfRead$IDFac <- factor(dfRead$ID,labels=c("æ","ɑ:","ɒ","e","ə","ɜ:","i:","ɪ","ɔ:","u:","ʊ","ʌ"))
NormalisedF1 <- dfRead$F1
NormalisedF2 <- dfRead$F2
#SVM classfication
set.seed(3)
index <- createDataPartition(dfRead,p = 0.60,list = FALSE)
head(index)
Test <- dfRead[index,]
Train <- dfRead[-index,]
svm_model <- svm(ID ~ F1+F2,data = dfRead,kernel = "radial")
pred = factor(predict(svm_model,dfRead))
dfRead$labelFac <- factor(dfRead$label,labels = c("æ","ʌ"))
plot(svm_model,Train,dfRead$ID ~ pred)
版权声明:本文内容由互联网用户自发贡献,该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容, 请发送邮件至 dio@foxmail.com 举报,一经查实,本站将立刻删除。