Kira (version 1.0.7)

brute.force: Brute force method for variable selection.

Description

Brute force method used to determine the smallest number of variables in a supervised classification model.

Usage

brute.force(func = NA, train, test, class.train,
            class.test,  args = NA, measure = "Rate Hits", 
            output = 10)

Value

best.model

Matrix with the names of the best combinations of variables, according to the evaluation measure used: accuracy, precision, recall etc.

text.model

Structure of the classification model used.

Arguments

func

Supervised classification function to be analyzed.

train

Data set of training, without classes.

test

Test data set.

class.train

Vector with training data class names.

class.test

Vector with test data class names.

args

Argument using in the classifier giving in 'func'.

measure

Measure to evaluate the model: "Rate Hits" (default), "Kappa Index", "Sensitivity", "Specificity", "Precision", "FP Rate", "FN Rate", "Negative Predictive Rate", "F-Score", "MCC", "ROC Are" or "PRC Area".
If measure = NA returns all metrics ordered by 'Rate Hits'.

output

Number of elements with the best combinations of variables in the matrix 'best.model' (default = 10).

Author

Paulo Cesar Ossani

Examples

data(iris) # data set

data  <- iris
names <- colnames(data)
colnames(data) <- c(names[1:4],"class")

#### Start - hold out validation method ####
dat.sample = sample(2, nrow(data), replace = TRUE, prob = c(0.7,0.3))
data.train = data[dat.sample == 1,] # training data set
data.test  = data[dat.sample == 2,] # test data set
class.train = as.factor(data.train$class) # class names of the training data set
class.test  = as.factor(data.test$class)  # class names of the test data set
#### End - hold out validation method ####

r <- (ncol(data) - 1)

res <- brute.force(func = "knn", train = data.train[,1:r], 
                   test = data.test[,1:r], class.train = class.train,  
                   class.test = class.test, args = "k = 1, dist = 'EUC'", 
                   measure = "Rate Hits", output = 20)
res$best.model
res$text.model

res <- brute.force(func = "regression", train = data.train[,1:r], 
                   test = data.test[,1:r], class.train = class.train, 
                   class.test = class.test, args = "intercept = TRUE", 
                   measure = "Rate Hits", output = 20)
res$best.model
res$text.model

test_a <- as.integer(rownames(data.test)) # test data index
class  <- data[,c(r+1)] # classes names
res <- brute.force(func = "lda", train = data[,1:r], test = test_a, 
                   class.train = class, class.test = class.test, 
                   args = "type = 'test', method = 'mle'", 
                   measure = "Rate Hits", output = 20)
res$best.model 
res$text.model