Novel evolutionary algorithms for supervised classification problems: an experimental study

Wang, Pu; Weise, Thomas; Chiong, Raymond

doi:10.1007/s12065-010-0047-7

Cited by 27 publications

(9 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…On the other hand, different neuro-evolutionary approaches have been successfully applied to a variety of benchmark problems and real-world classification tasks [25][26][27]32]. Our neuro-evolutionary algorithm, too, has been already tested and applied with success to several real-world problems, showing how such an approach can be useful in different classification problems, like automated trading strategy optimization [3,28], incipient fault diagnosis in electrical drives [29], automated diagnosis of skin diseases [30], etc.…”

Section: Neuro-evolutionary Classifiersmentioning

confidence: 99%

A Neuro-Evolutionary Approach to Electrocardiographic Signal Classification

Azzini

Dragoni

Tettamanzi

2014

Evolution, Complexity and Artificial Life

View full text Add to dashboard Cite

Abstract. This chapter presents an evolutionary ANN classifier system as a heartbeat classification algorithm designed according to the rules of the PhysioNet/Computing in Cardiology Challenge 2011 [7], whose aim is to develop an efficient algorithm able to run within a mobile phone, that can provide useful feedback in the process of acquiring a diagnostically useful 12-lead Electrocardiography (ECG) recording. The method used to solve this problem is to apply a very powerful natural computing analysis tool, namely evolutionary neural networks, based on the joint evolution of the topology and the connection weights relying on a novel similarity-based crossover. The chapter focuses on discerning between usable and unusable electrocardiograms tele-medically acquired from mobile embedded devices. A preprocessing algorithm based on the Discrete Fourier Trasform has been applied before the evolutionary approach in order to extract the ECG feature dataset in the frequency domain. Finally, a series of tests has been carried out in order to evaluate the performance and the accuracy of the classifier system for such a challenge.

show abstract

Section: Neuro-evolutionary Classifiersmentioning

confidence: 99%

A Neuro-Evolutionary Approach to Electrocardiographic Signal Classification

Azzini

Dragoni

Tettamanzi

2014

Evolution, Complexity and Artificial Life

View full text Add to dashboard Cite

show abstract

“…The review by Zhang [28], which provides a summary of the most important advances in classification with ANNs, makes it clear that the advantages of neural networks lie in different aspects: their capability to adapt themselves to the data without any explicit specification of functional or distributional form for the underlying model; they are universal functional approximators; they represent nonlinear and flexible solutions for modeling real world complex relationships; and, finally, they are able to provide a basis for establishing classification rules and performing statistical analysis. On the other hand, different neuro-evolutionary approaches have been successfully applied to a variety of benchmark problems and real-world classification tasks [29,30,31,32,33,3]. Our neuro-evolutionary algorithm, too, has been already tested and applied with success to several realworld problems, showing how such an approach can be useful in different classification problems, like financial time series modeling [34], automated trading strategy optimization [24,35], incipient fault diagnosis in electrical drives [36], automated diagnosis of skin diseases [37], brain-wave analysis [38], etc.…”

Section: Neuro-evolutionary Classifier Systemsmentioning

confidence: 99%

“…Artificial Neural Networks (ANNs) and Evolutionary Algorithms (EAs) are two important members of the family of computational methods collectively known as computational intelligence. Their advantages over conventional methods [1], like their conceptual and computational simplicity and their applicability to broad classes of optimization tasks, make them very attractive to approach those problems that pose difficulties to traditional techniques [2,3,4].…”

Section: Introductionmentioning

confidence: 99%

SimBa: A novel similarity-based crossover for neuro-evolution

Dragoni¹,

Azzini²,

Tettamanzi³

2014

Neurocomputing

View full text Add to dashboard Cite

International audienceThis work presents the SimBa (for Similarity-Based) crossover, a novel crossover operator specifically designed for the evolutionary optimization of neural network topologies that aims at overcoming one of the major problems generally related to the crossover operator, known as the permutation problem. The SimBa crossover starts by looking for a local similarity between two individuals selected from the population. The contribution of each neuron of the layer selected for the crossover is computed, and the neurons of each layer are reordered according to their contribution. Then, each neuron of the layer in the first individual is associated with the most similar neuron of the layer in the other individual, and the neurons of the layer of the second individual are re-ranked by considering the associations with the neurons of the first one. Finally, the neurons above a randomly selected cut-point are swapped to generate the offspring of the selected individuals. An approach exploiting this operator has been implemented and applied to six well-known benchmark classification problems. The experimental results, compared to those obtained by other techniques, show how this new crossover operator can help to produce compact neural networks with satisfactory generalization capability and accuracy

show abstract

“…Our GRD-XCS model is inspired by this methodology. Wang et al 16 used Information Gain as part of the fitness function in an EA (basically, the EA population contains several solutions and the fitness represents the appropriateness of each solution). They reported improved results when comparing their model to other machine learning algorithms.…”

Section: Related Workmentioning

confidence: 99%

Incorporating feature ranking and evolutionary methods for the classification of high-dimensional DNA microarray gene expression data

Abedini

Kirley

Chiong³

2013

AMJ

View full text Add to dashboard Cite

Our findings have shown that extracting feature quality information can assist the learning process and improve classification accuracy. On the other hand, relying exclusively on the feature quality information might potentially decrease the classification performance (e.g., using feature reduction). Therefore, we recommend a hybrid approach that uses feature quality information to direct the learning process by highlighting the more informative features, but at the same time not restricting the learning process to explore other features.

show abstract

Novel evolutionary algorithms for supervised classification problems: an experimental study

Cited by 27 publications

References 36 publications

A Neuro-Evolutionary Approach to Electrocardiographic Signal Classification

A Neuro-Evolutionary Approach to Electrocardiographic Signal Classification

SimBa: A novel similarity-based crossover for neuro-evolution

Incorporating feature ranking and evolutionary methods for the classification of high-dimensional DNA microarray gene expression data

Contact Info

Product

Resources

About