Empirical comparison of Dynamic Classifier Selection methods based on diversity and accuracy for building ensembles

Souto, Michael Vandesteen Silva; Soares, Rodrigo G. F.; Santana, Alixandre; Canuto, Anne M. P.

doi:10.1109/ijcnn.2008.4633992

Cited by 17 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many scholars of the accuracy of dynamic ensemble algorithm conducted in-depth study [3] - [7], and proposed a variety of dynamic ensemble strategy: Overall Local Accuracy, Local Class Accuracy, A Priori Selection and A Posteriori Selection.…”

Section: A Dynamic Ensemble Methodsmentioning

confidence: 99%

Dynamic combination of multiple classifiers based on central similarity

Wang

Liu

2011

The Fourth International Workshop on Advanced Computational Intelligence

View full text Add to dashboard Cite

According to the specific characteristics of samples, dynamic classifier ensemble chooses appropriate classifier for decision-making, which improve classification accuracy effectively, but increase the cost of running time. Therefore, Dynamic Combination of Multiple Classifiers Based on Central Similarity is proposed in this paper, which chooses different members classifier according to the similarity between classification samples and each class center to avoid validation process of neighborhood samples, and at the same time, adjust each corresponding weights to improve accuracy furthermore. The experiments demonstrate that this algorithm reduces the running time as well as improve the accuracy of integration classification, besides, choice of classifiers don't depend on neighborhood samples any more, so it shows a higher accuracy of classification for small scale sample training set.

show abstract

Section: A Dynamic Ensemble Methodsmentioning

confidence: 99%

Dynamic combination of multiple classifiers based on central similarity

Wang

Liu

2011

The Fourth International Workshop on Advanced Computational Intelligence

View full text Add to dashboard Cite

show abstract

“…As the clusters are fixed once for all, many different instances might share the same region of competence. In contrast, kNN methods give different regions of competence from one instance to another, which allows for more flexibility but at the expense of a higher computational cost [11].…”

Section: Evaluation and Selection Of The Best Classifiermentioning

confidence: 99%

Random Forest for Dissimilarity-Based Multi-View Learning

Bernard

Cao

Sabourin

2020

Handbook of Pattern Recognition and Computer Vision

View full text Add to dashboard Cite

Many classification problems are naturally multi-view in the sense their data are described through multiple heterogeneous descriptions. For such tasks, dissimilarity strategies are effective ways to make the different descriptions comparable and to easily merge them, by (i) building intermediate dissimilarity representations for each view and (ii) fusing these representations by averaging the dissimilarities over the views. In this work, we show that the Random Forest proximity measure can be used to build the dissimilarity representations, since this measure reflects similarities between features but also class membership. We then propose a Dynamic View Selection method to better combine the view-specific dissimilarity representations. This allows to take a decision, on each instance to predict, with only the most relevant views for that instance. Experiments are conducted on several real-world multi-view datasets, and show that the Dynamic View Selection offers a significant improvement in performance compared to the simple average combination and two state-ofthe-art static view combinations.

show abstract

“…by using a preselected committee of the individual classifier and making the final decision on the basis of a voting rule. 28 The random reference classifier used for dynamic selection was proposed in Ref. 29.…”

Section: Introductionmentioning

confidence: 99%

Improved Adaptive Splitting and Selection: The Hybrid Training Method of a Classifier Based on a Feature Space Partitioning

Jackowski

Krawczyk

Woźniak

2014

Int. J. Neur. Syst.

View full text Add to dashboard Cite

Currently, methods of combined classification are the focus of intense research. A properly designed group of combined classifiers exploiting knowledge gathered in a pool of elementary classifiers can successfully outperform a single classifier. There are two essential issues to consider when creating combined classifiers: how to establish the most comprehensive pool and how to design a fusion model that allows for taking full advantage of the collected knowledge. In this work, we address the issues and propose an AdaSS+, training algorithm dedicated for the compound classifier system that effectively exploits local specialization of the elementary classifiers. An effective training procedure consists of two phases. The first phase detects the classifier competencies and adjusts the respective fusion parameters. The second phase boosts classification accuracy by elevating the degree of local specialization. The quality of the proposed algorithms are evaluated on the basis of a wide range of computer experiments that show that AdaSS+ can outperform the original method and several reference classifiers.

show abstract

Empirical comparison of Dynamic Classifier Selection methods based on diversity and accuracy for building ensembles

Cited by 17 publications

References 17 publications

Dynamic combination of multiple classifiers based on central similarity

Dynamic combination of multiple classifiers based on central similarity

Random Forest for Dissimilarity-Based Multi-View Learning

Improved Adaptive Splitting and Selection: The Hybrid Training Method of a Classifier Based on a Feature Space Partitioning

Contact Info

Product

Resources

About