fork download
copy 
  1. # install.packages("rpart.plot", repos="http://c...content-available-to-author-only...o.com/", dependencies=TRUE)
  2. # install.packages('rpart', dependencies=TRUE)
  3. # install.packages('e1071', dependencies=TRUE)
  4. # install.packages('aod', dependencies = TRUE) 
  5. # install.packages('ggplot2', dep = TRUE)
  6. # install.packages('adabag',dependencies=T)
  7. # install.packages("ada",dep = T)
  8. # install.packages('randomForest', dependencies=TRUE)
  9. # install.packages("gbm")
  10.  
  11. library(rpart)
  12. library(rpart.plot)
  13. library(class)
  14. library(e1071)
  15. library(aod)
  16. library(mlbench)
  17. library(ggplot2)
  18. require(nnet)
  19. library(adabag)
  20. library(ada)
  21. library(randomForest)
  22. library(gbm)
  23.  
  24.  
  25. DataSetList <- c('http://a...content-available-to-author-only...i.edu/ml/machine-learning-databases/wine/wine.data',
  26.                  'http://a...content-available-to-author-only...i.edu/ml/machine-learning-databases/adult/adult.data',
  27.                  'http://a...content-available-to-author-only...i.edu/ml/machine-learning-databases/breast-cancer-wisconsin/wdbc.data',
  28.                  'http://a...content-available-to-author-only...i.edu/ml/machine-learning-databases/car/car.data',
  29.                  'http://a...content-available-to-author-only...i.edu/ml/machine-learning-databases/haberman/haberman.data')
  30.  
  31. classvar <- c(1,15,2,7,4)
  32.  
  33. for(j in 1:5)
  34.   {
  35.  
  36.   j=1
  37. d=read.csv(DataSetList[j],header = FALSE)
  38.  
  39. d[ is.na(d) ] <- 0
  40. d[d == "?"] <- 0
  41. apply(d,2,function(x) sum(is.na(x)))
  42.  
  43. set.seed(100)
  44.  
  45. bagging_accuracy <- 1:10
  46. adaboost_accuracy <- 1:10
  47. gboosting_accuracy <- 1:10
  48. forest_accuracy <- 1:10
  49. knn_accuracy <- 1:10
  50.  
  51.   d<-d[sample(nrow(d)),]
  52.   folds <- cut(seq(1,nrow(d)),breaks=10,labels=FALSE)
  53.   for(i in 1:10)
  54.   {
  55.     testIndexes <- which(folds==i,arr.ind=TRUE)
  56.     testData <- d[testIndexes, ]
  57.     trainingData <- d[-testIndexes, ]
  58.  
  59.     trainingData[trainingData == "?"] <- NA
  60.     trainingData<- trainingData[complete.cases(trainingData),]
  61.     testData[testData == "?"] <- NA
  62.     testData<- testData[complete.cases(testData), ]
  63.  
  64.     Class<-as.factor(trainingData[,classvar[j]])
  65.  
  66.     #-----------Bagging
  67.     model <- bagging(Class ~ ., data=trainingData[,-classvar[j]],coob=TRUE)
  68.     p=predict(model,testData[,-classvar[j]])
  69.     accuracy_bag <- sum(testData[,classvar[j]]==p)/length(p)*100
  70.     bagging_accuracy[i] <- accuracy_bag
  71.     #cat("Method = ", "Bagging",", accuracy= ", accuracy_bag,"\n")
  72.  
  73.  
  74.     #-----------ADA Boosting
  75.     model <- boosting(Class ~ ., data = trainingData[,-classvar[j]], boos=TRUE, mfinal=20,coeflearn='Breiman')
  76.     p=predict(model,testData[,-classvar[j]])
  77.     accuracy_boost <- sum(testData[,classvar[j]]==p)/length(p)*100
  78.     adaboosting_accuracy[i] <- accuracy_boost
  79.     #cat("Method = ", "Boosting",", accuracy= ", accuracy_boost,"\n") 
  80.  
  81.     #------------Gradient Boosting
  82.     gbm = gbm(Class ~ ., data=trainingData[,-classvar[j]], n.trees=1000,shrinkage=0.01,distribution="gaussian",interaction.depth=7,bag.fraction=0.9,cv.fold=10,n.minobsinnode = 50)
  83.     pred <- predict(gbm, newdata=testData[,-classvar[j]],OOB=TRUE, type = "response")
  84.     accuracy_gboost <- sum(testData[,classvar[j]]==round(pred))/length(pred)*100
  85.     gboosting_accuracy[i] <- accuracy_gboost
  86.    # cat("Method = ", "Gradient Boosting",", accuracy= ", accuracy_gboost,"\n")
  87.  
  88.     #-----------kNN----------
  89.     predicted_knn <- knn(trainingData[,-classvar[j]], testData[,-classvar[j]],Class, k =9)
  90.     accuracy_knn <- sum(predicted_knn == testData[,classvar[j]])/length(predicted_knn)*100
  91.     knn_accuracy[i] <- accuracy_knn
  92.    # cat("Method = ", "Knn",", accuracy= ", accuracy_knn,"\n")
  93.  
  94.     #-----------Random Forest
  95.     fit <- randomForest(Class ~ ., data=trainingData[,-classvar[j]], mtry=3,importance=TRUE, na.action=na.omit)
  96.     pred <- predict(fit, newdata=testData[,-classvar[j]],OOB=TRUE, type = "response")
  97.     accuracy_forest <- sum(testData[,classvar[j]]==pred)/length(pred)*100
  98.     forest_accuracy[i] <- accuracy_forest
  99.    # cat("Method = ", "Random Forest",", accuracy= ", accuracy_forest,"\n")
  100.  
  101.  
  102.     cat("\n")
  103.  
  104.  
  105.   }
  106.  
  107.   acc_forest <- mean(forest_accuracy)
  108.   acc_knn <- mean(knn_accuracy)
  109.   acc_gboost <- mean(gboosting_accuracy)
  110.   acc_boost <- mean(adaboost_accuracy)
  111.   #acc_bag <- mean(bagging_accuracy)
  112.  
  113.   cat("Total Accuracy-")
  114.   cat("Data Set : ",j,"\n")
  115.   cat("Method = ", "Random Forest",", accuracy= ", acc_forest,"\n")
  116.   cat("Method = ", "Knn",", accuracy= ", acc_knn,"\n")
  117.   cat("Method = ", "Gradient Boosting",", accuracy= ", acc_gboost,"\n")
  118.   cat("Method = ", "Ada Boosting",", accuracy= ", acc_boost,"\n") 
  119.   cat("Method = ", "Bagging",", accuracy= ", acc_bag,"\n")
  120.  
  121. }
  122.  
  123.  
  124.  
Success #stdin #stdout #stderr 0.33s 42348KB
comments (?)
stdin
copy 
Standard input is empty
stdout
copy 
Standard output is empty
stderr
copy 
Error in library(aod) : there is no package called ‘aod’
Execution halted
https://ideone.com/HMyX1s
language:
R (R 3.5.2)
created:
16 days ago
visibility:
public

Share or Embed source code

Discover > Sphere Engine API

The brand new service which powers Ideone!

Discover > IDE Widget

Widget for compiling and running the source code in a web browser!