How to use Different Algorithms using Caret Package in R

The caret package in R is a powerful tool for performing machine learning tasks, including training and evaluating models, feature selection, and hyperparameter tuning. It provides a unified interface to the various algorithms, making it easy to switch between different models and compare their performance. 

Caret package in R

The caret (Classification And REgression Training) package in R Programming Language is the comprehensive toolkit for training and evaluating machine learning models. It provides a unified interface for working with various algorithms, handling data preprocessing tasks, feature selection, model tuning, and performance evaluation.

Classification Algorithm using CARET Package

Now let’s train a random forest model on the iris dataset using the CARET Package.

R

library(caret)
data(iris)

# the model
model <- train(Species ~ ., data = iris,
               method = "rf")

# Evaluate the model
results <- trainControl(method = "cv",
                        number = 15)
performance <- train(Species ~ ., data = iris,
                     method = "rf",
                     trControl = results)

# Print
print(performance)

Output:

Random Forest 

150 samples
  4 predictor
  3 classes: 'setosa', 'versicolor', 'virginica' 

No pre-processing
Resampling: Cross-Validated (15 fold) 
Summary of sample sizes: 140, 140, 141, 140, 139, 140, ... 
Resampling results across tuning parameters:

  mtry  Accuracy   Kappa    
  2     0.9544108  0.9317932
  3     0.9470034  0.9206821
  4     0.9536700  0.9303397

Accuracy was used to select the optimal model using the largest value.
The final value used for the model was mtry = 2.

The below example is more or less the same as the above but the main difference lies in the method which you have used as a method. In the previous example, it was a random forest but in this it is rpart.

R

library(caret)
data(iris)

# the model
model <- train(Species ~ ., data = iris,
               method = "rpart")

# Evaluate the model
results <- trainControl(method = "cv", number = 12)
performance <- train(Species ~ .,
                     data = iris,
                     method = "rpart",
                     trControl = results)

# Print
print(performance)

Output:

CART 

150 samples
  4 predictor
  3 classes: 'setosa', 'versicolor', 'virginica' 

No pre-processing
Resampling: Cross-Validated (12 fold) 
Summary of sample sizes: 138, 137, 137, 138, 138, 138, ... 
Resampling results across tuning parameters:

  cp    Accuracy   Kappa    
  0.00  0.9203297  0.8804324
  0.44  0.7456502  0.6236446
  0.50  0.3977411  0.1111111

Accuracy was used to select the optimal model using the largest value.
The final value used for the model was cp = 0.

Regression Algorithm using Caret Package

Caret stands for the Classification and Regression Task also we have seen two examples of classification tasks above. Now let’s see how can we train a regression model using a linear method that is “lm” in the method parameter of the function.

R

library(caret)

data <- mtcars
train_control <- trainControl(method="cv",
                              number=4)

model_lm <- train(mpg~.,
                  data=data,
                  method="lm",
                  trControl=train_control)

print(model_lm)

Output:

Linear Regression 

32 samples
10 predictors

No pre-processing
Resampling: Cross-Validated (4 fold) 
Summary of sample sizes: 24, 24, 24, 24 
Resampling results:

  RMSE     Rsquared  MAE     
  4.04832  0.631639  3.300931

Tuning parameter 'intercept' was held constant at a value of TRUE