majority: Majority Classifier

Description Usage Arguments Details Value See Also Examples

View source: R/majority.R

Description

A classifier that always predicts the class with the highest weighted prior probability.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
majority(x, ...)

## S3 method for class 'formula'
majority(formula, data, weights = rep(1, nrow(data)), ...,
  subset, na.action)

## S3 method for class 'data.frame'
majority(x, ...)

## S3 method for class 'matrix'
majority(x, grouping, weights = rep(1, nrow(x)), ..., subset,
  na.action = na.fail)

## Default S3 method:
majority(x, grouping, weights = rep(1, nrow(x)), ...)

Arguments

x

(Required if no formula is given as principal argument.) A matrix or data.frame or Matrix containing the explanatory variables.

formula

A formula of the form groups ~ x1 + x2 + ..., that is, the response is the grouping factor and the right hand side specifies the discriminators.

data

A data.frame from which variables specified in formula are to be taken.

weights

Observation weights to be used in the fitting process, must be non-negative.

subset

An index vector specifying the cases to be used in the training sample. (NOTE: If given, this argument must be named.)

na.action

A function to specify the action to be taken if NA's are found. The default action is first the na.action setting of options and second na.fail if that is unset. An alternative is na.omit, which leads to rejection of cases with missing values on any required variable. (NOTE: If given, this argument must be named.)

grouping

(Required if no formula is given as principal argument.) A factor specifying the class membership for each observation.

...

Further arguments.

Details

The majority classifier usually serves to determine a baseline for more sophisticated classification methods. This function is also a helper function needed to combine mixture models and recursive partitioning with a majority classifier. The weighted prior probabilities are calculated as

p_g = ∑_{n:y_n=g} w_n/(∑_n w_n)

Value

An object of class "majority", a list containing the following components:

prior

Weighted class prior probabilities.

counts

The number of observations per class.

lev

The class labels (levels of grouping).

N

The number of observations.

weights

The observation weights used in the fitting process.

predictors

The names of the predictor variables.

call

The (matched) function call.

See Also

Other majority: majorityGenerative, predict.majorityGenerative, predict.majority

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
library(mlbench)
data(PimaIndiansDiabetes)

train <- sample(nrow(PimaIndiansDiabetes), 500)

# weighting observations from classes pos and neg according to their
# frequency in the data set:
ws <- as.numeric(1/table(PimaIndiansDiabetes$diabetes)
    [PimaIndiansDiabetes$diabetes])

fit <- majority(diabetes ~ ., data = PimaIndiansDiabetes, weights = ws,
    subset = train)
pred <- predict(fit, newdata = PimaIndiansDiabetes[-train,])
mean(pred$class != PimaIndiansDiabetes$diabetes[-train])

schiffner/locClass documentation built on May 29, 2019, 3:39 p.m.