Theoretical and Empirical Analysis of ReliefF and RReliefF

By Marko Robnik-Sikonja, Igor Kononenko. Machine-mediated learning, 2003

2,984 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Information Systems

View Paper: https://www.semanticscholar.org/paper/d398ad942320c1c2cf64da23ba519870f9f1d67d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Information Systems

Citations in the corpus

Filter Methods for Feature Selection - A Comparative Study (2007, neutral): “6 (??) 100 200 150 4,6,8,11,17,19 [1-11],[13-20],38,44,100, 112,120,123,144,172,180,185,191,197,200”
A Strategy for Short-Term Load Forecasting by Support Vector Regression Machines (2013, methodology): “, feature selection based on genetic algorithms and information theory (GA&IT) [44], feature selection using mutual information and maximum-relevance minimum-redundancy criterion (MRMR) [45], feature selection using ReliefF algorithm [46], sequential feature selection with stepwise regression (Steep”
Trace Ratio Criterion for Feature Selection (2008, methodology): “According to the strategy of utilizing class label information, feature selection algorithms can be roughly divided into three categories, namely unsupervised feature selection (Dy & Brodley 2004), semisupervised feature selection (Zhao & Liu 2007a), and supervised feature selection ( Robnik-Sikonja”
Spatially Uniform ReliefF (SURF) for computationally-efficient filtering of gene-gene interactions (2009, methodology): “With Regression ReliefF (RReliefF), ReliefF is broadened to handle continuous attributes and endpoints [23,24].”
Feature Selection for Classification: A Review (2014, methodology): “In [56], the authors related the relevance evaluation criterion of ReliefF to the hypothesis of margin maximization, which explains why the algorithm provide superior performance in many applications.”
Machine Learning and Data Mining Methods in Diabetes Research (2017, methodology): “[28] applied Random Forest (RF) [35] and RReliefF [36] to evaluate a number of features, with respect to their ability to predict the short-term subcutaneous glucose concent rations.”
Nearest neighbor imputation algorithms: a critical evaluation (2016, methodology): “As previously shown in classification problems [17], this number is context-dependent, however the choice of the most parsimonious model seems reasonable as we do not expect many correlations among variables.”
Measuring the Effectiveness of Various Features for Thematic Information Extraction From Very High Resolution Remote Sensing Imagery (2015, methodology): “ReliefF evaluates the relevance of features by feature weighting according to how well features distinguish samples [23].”
Differential evolution for filter feature selection based on information theory and feature ranking (2018, methodology): “To achieve this goal, a novel ﬁlter evaluation criterion (named MIRFFS) based on the concepts of mutual relevance, RelifF [27] and Fisher Score [28] is proposed, and using this proposed criterion, the standard DE and multi-objective DE (MODE) based feature selection approaches are developed.”
Searching for Interacting Features (2007, methodology): “Representative measures used for evaluating relevance include: distance measures [Kononenko, 1994; Robnik-Sikonja and Kononenko, 2003], information measures [Fleuret, 2004], and consistency measures [Almuallim and Dietterich, 1994], to name a few.”