Combating the Small Sample Class Imbalance Problem Using Feature Selection

By Michael Wasikowski, Xue-wen Chen. IEEE Transactions on Knowledge and Data Engineering, 2010

394 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/bdc626bcbcb792df53debe26db9a5b09f6a0e90f

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 Electrical and Electronic Engineering

Citations in the corpus

A Survey of Predictive Modeling on Imbalanced Domains (2016, background): “Feature selection is recommended [Wasikowski and Chen 2010] and is also pointed as the solution for addressing the class imbalance problem [Mladenic and Grobelnik 1999; Zheng et al. 2004; Chen and Wasikowski 2008; Van Der Putten and Van Someren 2004; Forman 2003].”
An insight into classification with imbalanced data: Empirical results and current trends on using data intrinsic characteristics (2013, background): “The combination of imbalanced data and the small sample size problem presents a new challenge to the research community [133].”
Feature selection methods for big data bioinformatics: A survey from the search perspective. (2016, background): “Wasikowski and Chen [158] carried out systematic comparisons of three types of methods developed for imbalanced data classification problems and seven FS metrics evaluated on small sample data sets from different applications.”
Class Imbalance Problem in Data Mining Review (2013, methodology): “Sampling techniques and algorithmic methods may not be enough to solve high dimensional class imbalance problems [5].”
Analysing the classification of imbalanced data-sets with multiple classes: Binarization techniques and ad-hoc approaches (2013, background): “Speciﬁcally, we must stress the signiﬁcance of several factors such as overlapping between classes [38], lack of representative data [59], small disjuncts [60,45], noisy data [49], dataset shift [43] and other issues which have interdependent effects with data distribution (imbalance).”
On the use of MapReduce for imbalanced big data using Random Forest (2014, background): “The presence of small sample size problems [54] which are related to the lack of density is also analysed.”
Dual autoencoders features for imbalance classification problem (2016, background): “the majority and the minority classes yields a good classification for both classes [2].”
Classification with class imbalance problem: A review (2015, methodology): “[66] CFS; correlation feature selection Helpful in alleviating class overlapping problem Extra computational cost due to included preprocessing task”
SMOTE for Learning from Imbalanced Data: Progress and Challenges, Marking the 15-year Anniversary (2018, background): “Therefore, learning classifiers cannot carry out a good generalization when there is not enough data to represent the boundaries of the problem (Jo & Japkowicz, 2004; Wasikowski & Chen, 2010).”
Feature selection for high-dimensional class-imbalanced data sets using Support Vector Machines (2014, methodology): “Filter and resampling methods have been used together with positive results in imbalanced data sets [5,29,35,40,43].”