Wine Quality Dataset Prediction Analysis using R and caret

winequality.R

packages <- c('caret')
if (length(setdiff(packages, rownames(installed.packages()))) > 0) {
  install.packages(setdiff(packages, rownames(installed.packages())))  
}
library(caret)

# Download dataset, if it does not exist.
fileName <- 'winequality-red.csv';
if (!file.exists(fileName)) {
  download.file(paste0('https://archive.ics.uci.edu/ml/machine-learning-databases/wine-quality/', fileName), fileName, method="curl")
}
fileName <- 'winequality-white.csv';
if (!file.exists(fileName)) {
  download.file(paste0('https://archive.ics.uci.edu/ml/machine-learning-databases/wine-quality/', fileName), fileName, method="curl")
}

data <- read.csv('winequality-white.csv', sep=';')
data <- rbind(data, read.csv('winequality-red.csv', sep=';'))

partition <- createDataPartition(data$quality, p = 0.75)[[1]]
train <- data[partition,]
test <- data[-partition,]

# 0.5172 accuracy
#fit <- train(quality ~ ., data = train, method = 'plsRglm')

# 0.5172
#fit <- train(quality ~ ., data = train, method = 'plsRglm', preProcess = c("center", "scale"))

# 0.5209
fit <- train(quality ~ alcohol, data = train, method = 'plsRglm')

results <- round(predict(fit, newdata=test))
confusionMatrix(results, test$quality)

xoutput.txt

          Reference
Prediction   3   4   5   6   7   8   9
         3   0   0   0   0   0   0   0
         4   0   0   0   0   0   0   0
         5   1  14 272 169  22   7   0
         6   9  32 266 511 212  32   1
         7   0   0   2  29  32  12   1
         8   0   0   0   0   0   0   0
         9   0   0   0   0   0   0   0

Overall Statistics
                                          
               Accuracy : 0.5018          
                 95% CI : (0.4772, 0.5265)
    No Information Rate : 0.4366          
    P-Value [Acc > NIR] : 7.299e-08