Optimal construction of one-against-one classifier based on meta-learning

Kang, Seokho; Cho, Sungzoon

doi:10.1016/j.neucom.2015.04.048

Cited by 11 publications

(2 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, meta-learning has been used to tune the parameters of a classifier [18], [40]. Other works include the use of meta-learning in other contexts, such as the treatment of noisy data [16], [55], instance selection [35], classifier building [31] and even the clustering field [13], [62].…”

Section: Introductionmentioning

confidence: 99%

A Meta-Learning Recommendation System for Characterizing Unsupervised Problems: On Using Quality Indices to Describe Data Conformations

Sáez

Corchado

2019

IEEE Access

View full text Add to dashboard Cite

The clustering of a new unsupervised problem usually requires knowing both if the samples may be separable in different groups and the number of these groups. This information, which has a great impact on the results obtained, is generally unknown beforehand. A wide explored research line in the literature proposes to use some metrics, known as quality indices, to determine the number of clusters in a dataset. However, they may lead to variable results depending on the metric chosen. This research analyzes the usage of a novel meta-learning system for determining the number of clusters in unsupervised data, called Meta-Learning Recommendation System for Cluster Cardinality Estimation (MLRS-CCE). It is based on the idea of using quality metrics not as a solution to the problem, but as a means to characterize the inner structure of each dataset and employing this information to detect when unsupervised data is not uniform and suggest additional information about the number of clusters in the data. In order to achieve such goals a large collection of both real-world and synthetic datasets, in which the number of clusters is known a priori, are used to build the system and check its performance. The meta-learning system was successfully tested on such data, showing that it is accurate enough, both separating uniform data from non-uniform one and predicting cluster cardinality when it is compared to the results given by individual quality indices. INDEX TERMS Cluster cardinality, data uniformity, meta-learning, quality indices, unsupervised learning.

show abstract

Section: Introductionmentioning

confidence: 99%

A Meta-Learning Recommendation System for Characterizing Unsupervised Problems: On Using Quality Indices to Describe Data Conformations

Sáez

Corchado

2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Originally, RVM was designed to fulfill binary classification and univariate regression. For multi-classification problems, although several approaches have been developed to construct a multiclass RVM among which two methods are suggested in literature containing one against all (OAA) and one against one (OAO) [20], [21] they are inconvenient and time-consuming [20], [22]. A. Thayananthan [23] proposed a multivariate relevance vector machine (MRVM) extending from RVM which was originally used to resolve the pose estimation problem.…”

Section: Introductionmentioning

confidence: 99%

Multiple-Kernel MRVM With LBFO Algorithm for Fault Diagnosis of Broken Rotor Bar in Induction Motor

Zhao

Wang

2019

IEEE Access

View full text Add to dashboard Cite

Induction motors are key equipments widely used in modern industries. Fault diagnosis of broken rotor bar (BRB) timely and accurately is very important to ensure the reliable operation of induction motors (IMs). In this study, a multivariate relevance vector machine with multiple Gaussian kernels (MKMRVM) and principal component analysis (PCA) are developed to construct a classification model. Then an improved bacterial foraging optimization combining with Levy fight mechanism, named LBFO, is employed to tune the kernel parameters of MKMRVM to obtain the optimal fault diagnosis model. Finally, The LBFO-based MKMRVM classification model is used to identify the diagnosis of BRB of IMs. The performance is assessed based on a comprehensive experiment of fault diagnosis of BRB. The experimental results demonstrate that the proposed method may achieve high diagnosis accuracy for different noise levels of diagnosis signals and is superior to other related fault diagnosis techniques.

show abstract

Selection of the Best Base Classifier in One-Versus-One Using Data Complexity Measures

Morán-Fernández

Bolón-Canedo

Alonso‐Betanzos

2016

Advances in Artificial Intelligence

View full text Add to dashboard Cite

Optimal construction of one-against-one classifier based on meta-learning

Cited by 11 publications

References 46 publications

A Meta-Learning Recommendation System for Characterizing Unsupervised Problems: On Using Quality Indices to Describe Data Conformations

A Meta-Learning Recommendation System for Characterizing Unsupervised Problems: On Using Quality Indices to Describe Data Conformations

Multiple-Kernel MRVM With LBFO Algorithm for Fault Diagnosis of Broken Rotor Bar in Induction Motor

Selection of the Best Base Classifier in One-Versus-One Using Data Complexity Measures

Contact Info

Product

Resources

About