The document evaluates the quality of red wine from the 'vinho verde' region using a dataset with physicochemical and sensory variables. Three methods were applied for analysis: CART, bagging, and random forest, with random forest proving to be the most effective in minimizing test error rates. Key attributes influencing wine quality include alcohol, total sulfur dioxide, volatile acidity, and sulfates.

1. Red Wine
Quality
Evaluation
Weiyang Bi
Shilin Wang
Zheng Xue

2. Data Description
• Source:
Paulo Cortez, University of Minho, Guimarães,
Portugal, http://www3.dsi.uminho.pt/pcortez A.
Cerdeira, F. Almeida, T. Matos and J. Reis,
Viticulture Commission of the Vinho Verde
Region(CVRVV), Porto, Portugal @2009

3. Data Description
• The dataset is related to red variant of
the Portuguese "Vinho Verde" wine.
• Due to privacy and logistic issues, only
physicochemical (inputs) and sensory
(the output) variables are available.

4. Dataset
> nrow(data[!complete.cases(data),])
[1] 0
Missing values check

5. Attribute information
Input variables (based on physicochemical
tests): 1 - fixed acidity 2 - volatile acidity 3 - citric
acid 4 - residual sugar 5 - chlorides 6 - free sulfur
dioxide 7 - total sulfur dioxide 8 - density 9 - pH 10 -
sulphates 11 - alcohol Output variable (based on
sensory data): 12 - quality (score between 0 and 10)

6. Correlation matrix

7. R code for training set
and test set
B<-20
for(i in 1:B){
set.seed(i)
indexes<-sample(1:nrow(data),size=1000,replace=F)
train<-data[indexes[1:1000],]
test<-data[-indexes[1:1000],]
}

8. Methods
Three methods were applied to the data set:
1) CART
2) Bagging
3) Random Forest

9. Classification and
Regression Trees (CART)

10. CP Table

11. Pruned Tree

13. Variable Selection
Total
sulfur
dioxide
Volatile
acidity
sulfates
Residual
sugar
alcohol

14. Number of splits

15. Error rate

16. Bagging

17. Merging data

18. CP Table

19. Misclassification Rate

20. Misclassification rate of
bagged 100 trees

21. ROC Graph
0.0 0.2 0.4 0.6 0.8 1.0
0.00.20.40.60.81.0
1-Sepecificity
Sensitivity
Best single tree:0.64
Bagged 100 trees:0.644

22. Frequency Table

23. Evaluation of Variable Importance

24. Random Forest

25. Data Structure

26. Random Forest Fit

27. Random Forest Plot

28. Importance

29. Relative Variable Importance

30. Partial Dependence Plot
Alcohol Sulphates
Volatiles acidity Total sulfur dioxide
Partialdependence
Volatiles acidity
Partialdependence
Partialdependence
Partialdependence

31. CART Bagging and RF Comparison
CART Bagging Random Forest
Variable
Selection
Alcohol
total sulfur dioxide
volatile acidity
Sulphates
Residual sugar
Alcohol
Sulphates
volatile acidity
total sulfur dioxide
Density
fixed acidity
residual sugar
citric acid
pH
free sulfur dioxide
Chlorides
Alcohol
Sulphates
volatile acidity
total sulfur dioxide
Density
Chlorides
fixed acidity
free sulfur dioxide
pH
citric acid
residual sugar

32. CART Bagging and RF Comparison

33. Conclusion
• Random forest is the best prediction
tool in this case over CART and bagging
in terms of the lowest estimate test
error rate.