ReliefFAttributeEval :
Evaluates the worth of an attribute by repeatedly sampling an instance and considering the value of the given attribute for the nearest instance of the same and different class. Can operate on both discrete and continuous class data.
For more information see:
Kenji Kira, Larry A. Rendell: A Practical Approach to Feature Selection. In: Ninth International Workshop on Machine Learning, 249-256, 1992.
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.
BibTeX:
@inproceedings{Kira1992, author = {Kenji Kira and Larry A. Rendell}, booktitle = {Ninth International Workshop on Machine Learning}, editor = {Derek H. Sleeman and Peter Edwards}, pages = {249-256}, publisher = {Morgan Kaufmann}, title = {A Practical Approach to Feature Selection}, year = {1992} } @inproceedings{Kononenko1994, author = {Igor Kononenko}, booktitle = {European Conference on Machine Learning}, editor = {Francesco Bergadano and Luc De Raedt}, pages = {171-182}, publisher = {Springer}, title = {Estimating Attributes: Analysis and Extensions of RELIEF}, year = {1994} } @inproceedings{Robnik-Sikonja1997, author = {Marko Robnik-Sikonja and Igor Kononenko}, booktitle = {Fourteenth International Conference on Machine Learning}, editor = {Douglas H. Fisher}, pages = {296-304}, publisher = {Morgan Kaufmann}, title = {An adaptation of Relief for attribute estimation in regression}, year = {1997} } Valid options are:
-M <num instances> Specify the number of instances to sample when estimating attributes. If not specified, then all instances will be used.
-D <seed> Seed for randomly sampling instances. (Default = 1)
-K <number of neighbours> Number of nearest neighbours (k) used to estimate attribute relevances (Default = 10).
-W Weight nearest neighbours by distance
-A <num> Specify sigma value (used in an exp function to control how quickly weights for more distant instances decrease. Use in conjunction with -W. Sensible value=1/5 to 1/10 of the number of nearest neighbours. (Default = 2)
@author Mark Hall (mhall@cs.waikato.ac.nz)
@version $Revision: 5511 $