jranaraki / DependencyDegree

This is an implementation of Fuzzy Rough Dependency Degree (FRDD) to calculate the importance of selected feautres

title	author	date
DependencyDegree	Javad Rahimipour Anaraki	23/05/18

Use case

To determine fuzzy rough dependency degree of features based on New Approaches to Fuzzy-Rough Feature Selection which can be used as a fitness function of evolutionary algorithms

Compile

This code can be run using MATLAB R2006a and above

Run

To run the code, call dpendency.m function with two inputs, address to a dataset and a binary vector with size of features of the dataset. For instance, calling the function with dependency('Data/wine.csv', [0 0 0 1 0 1 0 1 1 0 0 0 0]) would calculate fuzzy rough dependency degree of selected features {4, 6, 8 ,9} for wine.csv dataset.

Datasets can be downloaded from UCI Machine Learning Repository

Note

Datasets should have no column and/or row names, and the class values should be all numeric