CoXCS: A Coevolutionary Learning Classifier Based on Feature Space Partitioning

Abedini, Mani; Kirley, Michael

doi:10.1007/978-3-642-10439-8_37

Cited by 14 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the literature parallel working Learning Classifier Systems are typically used to improve the algorithmic performance of a classifier system or in order to be able to deal with distributed data [25]. For example, some authors apply multiple coevolving Michigan-style Learning Classifier Systems on decompositions of the problem space to increase the solution quality on large problems [1,10,19,30]. Moreover, coevolving Michigan-style Learning Classifier Systems have been used for social modelling on the El' Farol problem [13].…”

Section: Related Workmentioning

confidence: 98%

Learning classifier systems to evolve classification rules for systems of memory constrained components

Scheidler

Middendorf²

2011

Evol. Intel.

View full text Add to dashboard Cite

In this paper we study how to solve classification problems in computing systems that consist of distributed, memory constrained components. Interacting Pittsburgh-style Learning Classifier Systems are used to generate sets of classification rules that can be deployed on the components. We show that this approach distributes the knowledge and enables the components to solve complex classification problems in cooperation. We study the structure and properties of the evolved rule sets and analyse the way the components share their knowledge. Moreover, we investigate the influence of different communication topologies and the introduction of communication costs on the emerging patterns of cooperation and on the classification performance of the whole system.

show abstract

Section: Related Workmentioning

confidence: 98%

Learning classifier systems to evolve classification rules for systems of memory constrained components

Scheidler

Middendorf²

2011

Evol. Intel.

View full text Add to dashboard Cite

show abstract

“…CoXCS is a coevolutionary learning classifier based on feature space partitioning [2]. It extends the XCS model by introducing a coevolutionary approach.…”

Section: A Coxcsmentioning

confidence: 99%

Gene Expression Classification with a Novel Coevolutionary Based Learning Classifier System on Public Clouds

Vecchiola

Abedini

Kirley

et al. 2010

2010 Sixth IEEE International Conference on E-Science Workshops

Self Cite

View full text Add to dashboard Cite

Microarray technology allows for the simultaneous monitoring of thousands of genes expressions per sample. Unfortunately, the classification of these samples into distinct classes is often difficult as the number of genes (features) greatly exceeds the number of samples. Consequently, there is a need to investigate new, robust machine learning techniques capable of accurately classifying microarray data. In this paper, we present a coevolutionary learning classifier system based on feature set partitioning to classify gene expression data. A distributed implementation, which leverages Cloud computing technologies, is used to address the inherent computational costs of our model. The development and execution of this application was done using the Aneka middleware on the public Cloud (Amazon EC2) infrastructure. Experiments conducted using gene expression profiles demonstrates that the proposed implementation outperforms other well-known classifiers in terms of accuracy. Preliminary analysis into the impact of different Cloud setups on the performance of the classifier are also reported.

show abstract

“…It was shown in Tumer and Oza (2003) that feature partitioning potentially helps in overcoming the curse of dimensionality and avoids the feature selection drawbacks. There are different popular strategies for creating feature subset-based ensembles, such as manual selection (Gershoff and Schulenburg 2007), feature bagging (or random subspace) (Ho 1998), equal linear division of the feature space (Abedini and Kirley 2009), or by employing feature selection algorithms (Debie et al 2013a). The random subspace approach in particular has been studied extensively in the literature, especially its application to the decision tree classifier (random forests) (Breiman 2001).…”

Section: A Generic Framework For Categorizing Ensemblesmentioning

confidence: 99%

“…Individuals were used to classify the partially masked training data corresponding to the feature in focus. A co‐evolutionary multi‐population XCS was proposed (Abedini and Kirley ). Co‐evolutionary multi‐population XCS is based on a collection of independent populations of classifiers that are trained using different partitions of the feature space within the training data set.…”

Section: Introductionmentioning

confidence: 99%

On Taxonomy and Evaluation of Feature Selection‐Based Learning Classifier System Ensemble Approaches for Data Mining Problems

Debie

Shafi

Kasmarik

et al. 2016

Computational Intelligence

View full text Add to dashboard Cite

Ensemble methods aim at combining multiple learning machines to improve the efficacy in a learning task in terms of prediction accuracy, scalability, and other measures. These methods have been applied to evolutionary machine learning techniques including learning classifier systems (LCSs). In this article, we first propose a conceptual framework that allows us to appropriately categorize ensemble‐based methods for fair comparison and highlights the gaps in the corresponding literature. The framework is generic and consists of three sequential stages: a pre‐gate stage concerned with data preparation; the member stage to account for the types of learning machines used to build the ensemble; and a post‐gate stage concerned with the methods to combine ensemble output. A taxonomy of LCSs‐based ensembles is then presented using this framework. The article then focuses on comparing LCS ensembles that use feature selection in the pre‐gate stage. An evaluation methodology is proposed to systematically analyze the performance of these methods. Specifically, random feature sampling and rough set feature selection‐based LCS ensemble methods are compared. Experimental results show that the rough set‐based approach performs significantly better than the random subspace method in terms of classification accuracy in problems with high numbers of irrelevant features. The performance of the two approaches are comparable in problems with high numbers of redundant features.

show abstract

CoXCS: A Coevolutionary Learning Classifier Based on Feature Space Partitioning

Cited by 14 publications

References 17 publications

Learning classifier systems to evolve classification rules for systems of memory constrained components

Learning classifier systems to evolve classification rules for systems of memory constrained components

Gene Expression Classification with a Novel Coevolutionary Based Learning Classifier System on Public Clouds

On Taxonomy and Evaluation of Feature Selection‐Based Learning Classifier System Ensemble Approaches for Data Mining Problems

Contact Info

Product

Resources

About