FSelector (version 0.34)

relief: RReliefF filter

Description

The algorithm finds weights of continous and discrete attributes basing on a distance between instances.

Usage

relief(formula, data, neighbours.count = 5, sample.size = 10)

Value

a data.frame containing the worth of attributes in the first column and their names as row names

Arguments

formula

a symbolic description of a model

data

data to process

neighbours.count

number of neighbours to find for every sampled instance

sample.size

number of instances to sample

Author

Piotr Romanski

Details

The algorithm samples instances and finds their nearest hits and misses. Considering that result, it evaluates weights of attributes.

References

• -Igor Kononenko: Estimating Attributes: Analysis and Extensions of RELIEF. In: European Conference on Machine Learning, 171-182, 1994.

• -Marko Robnik-Sikonja, Igor Kononenko: An adaptation of Relief for attribute estimation in regression. In: Fourteenth International Conference on Machine Learning, 296-304, 1997.

Examples

  data(iris)
  
  weights <- relief(Species~., iris, neighbours.count = 5, sample.size = 20)
  print(weights)
  subset <- cutoff.k(weights, 2)
  f <- as.simple.formula(subset, "Species")
  print(f)