R 지도학습 알고리즘_로지스틱

###지도학습 알고리즘 ##서포트 벡터 머신 auto=read.csv('autoparts.csv',header=T) auto1=auto[auto$prod_no=='90784-76001',c(2:11)] auto2=auto1[auto1$c_thickness<1000,] auto2$y_falty=ifelse((auto2$c_thickness<20)|(auto2$c_thickness>32),1,0) head(auto2) t_index=sample(1:nrow(auto2),size=nrow(auto2)*0.7) train=auto2[t_index,] test=auto2[-t_index,] nrow(train);nrow(test) install.packages('e1071') library(e1071) tune.svm(factor(y_falty)~fix_time+a_speed+b_speed+separation+s_separation+rate_terms+mpa+load_time+            highpressure_time,data=auto2,gamma=2^(-1:1),cost = 2^(2:4)) #훈련데이터 모델로 만든다 m=svm(factor(y_falty)~fix_time+a_speed+b_speed+separation+s_separation+rate_terms+mpa+load_time+         highpressure_time,data=auto2,gamma=1,cost =16) yhat_test=predict(m,test) table=table(reas=test$y_falty,predict(yhat_test)) table (table[1,1]+table[2,2])/sum(table) #정분류율 l=svm(factor(y_falty)~fix_time+a_speed+b_speed+separation+s_separation+rate_terms+mpa+load_time+         highpressure_time,data=train,gamma=1,cost =16,kernel='linear') yhat_test=predict(m,test) table=table(real=test$y_faulty,predict=yhat_test);table (table[1,1]+table[2,2])/sum(table) b=svm(factor(y_falty)~fix_time+a_speed+b_speed+separation+s_separation+rate_terms+mpa+load_time+         highpressure_time,data=train) yhat_test=predict(m,test) table=table(real=test$y_faulty,predict=yhat_test);table (table[1,1]+table[2,2])/sum(table) ##ROC커브 m=svm(factor(y_falty)~fix_time+a_speed+b_speed+separation+s_separation+rate_terms+mpa+load_time+         highpressure_time,data=train,gamma=2,cost=16) yhat_test=predict(m,test) library(Epi) install.packages('Epi') ROC(test=yhat_test,stat=test$y_falty,plot="ROC",AUC=T,main='SVM') lhat_test=predict(l,test) ROC(test=lhat_test,stat=test$y_falty,plot="ROC",AUC=T,main='SVM') bhat_test=predict(b,test) ROC(test=bhat_test,stat=test$y_falty,plot="ROC",AUC=T,main='SVM') ##예측 new.data=data.frame(fix_time=87,a_speed=0.609,b_speed=1.715,separation=242.7,                     s_separation=657.5,rate_terms=95,mpa=78,load_time=18.1,                     highpressure_time=82) predict(m,newdata=new.data) new.data=data.frame(fix_time=c(87,85.6),a_speed=c(0.609,0.472),b_speed=c(1.715,1.685),separation=c(242.7,243.3),                     s_separation=c(657.5,657.9),rate_terms=c(95,95),mpa=c(78,28.8),                     load_time=c(18.1,18.2),highpressure_time=c(82,60)) predict(m,newdata=new.data) ###서포트 벡터 회귀분석 auto=read.csv('autoparts.csv') auto1=auto[auto$prod_no=='90784-76991'] auto2=auto1[auto1$c_thinkness<1000,] auto2$y_falty=ifelse((auto2$c_thickness<20)|(auto2$c_thickness>21),1,0) t_index=sample(1:nrow(auto2),size=nrow(auto2*0.7)) train=auto2[t_index,] test=auto2[-t_index,] m=svm(c_thickness~fix_time+a_speed+b_speed+separation+s_separation+         rate_terms+mpa+load_time+highpressure_time,data=train,       gamma=1,cost=16) yhat_test=predict(m,test) plot(x=test$c_thickness,y=yhat_test,main='SVR') m2=lm(c_thickness~fix_time+a_speed+b_speed+separation+s_separation+rate_term+         mpa+load_time+highpressure_time,data=train) regression=read.csv('regression.csv',header = T) plot(regression$x,regression$y,pch=16,xlab='x',ylab='y') #LM m1=lm(y~x,data=regression) p1=predict(m1,newdata=regression) points(regression$x,p1,col='red',pch='L') #SVR m2=   svm(y~x,regression) p2=predict(m2,newdata=regression) ## 로지스틱회귀분석 auto=read.csv('autoparts.csv',header=T) auto1=auto[auto$prod_no=='90784-76001',c(2:11)] auto2=auto1[auto1$c_thickness<1000,] auto2$y_faulty=ifelse((auto2$c_thickness<20)|(auto2$c_thickness>32),1,0) auto2$y_faulty=as.factor(auto2$y_faulty) table(auto2$y_faulty) m=glm(y_faulty~fix_time+a_speed+b_speed+separation+         s_separation+rate_terms+mpa+load_time+         highpressure_time,data=auto2,family=binomial(logit)) #훈련데이터로만 모델 만들기 t_index=sample(1:nrow(auto2),size=nrow(auto2*0.7)) train=auto2[t_index,] test=auto2[-t_index,] m=glm(y_faulty~fix_time+a_speed+b_speed+separation+         s_separation+rate_terms+mpa+load_time+         highpressure_time,data=train,family=binomial(logit)) #승산비 기준값 설정하기 yhat=ifelse(m$fitted.values>=0.5,1,0) yhat table=table(real=train$y_faulty,predict=yhat) table yhat_test=p redict(m,test,type='response') ##다항 로지스틱스 모형 구현 : 범주가 3개 이상일 때 auto2$g_class=as.factor(ifelse(auto2$c_thickness<20,1,ifelse(auto2$c_thickness<32,2,3))) table(auto2$g_class) t_index=sample(1:nrow(auto2),size=nrow(auto2)*0.7) train=auto2[t_index,] test=auto2[-t_index,] install.packages('nnet') library(nnet) m=multinom(g_class~fix_time+a_speed+b_speed+separation+              s_separation+rate_terms+mpa+load_time+              highpressure_time,data=train) summary(m) head(m$fitted.values) yhat_test=predict(m,test) table=table(real=test$g_class,predict=yhat_test) table #정오표 (table[1,1]+table[2,2]+table[3,3])/sum(table)