Fuzzy relevance vector machine for learning from unbalanced data and noise

Li, Ding-Fang; Hu, Wenyi; Xiong, Wei; Yang, Jinbo

doi:10.1016/j.patrec.2008.01.009

Cited by 24 publications

(14 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In order to detect outliers, robust based methods are the most commonly used by researchers managing multivariate datasets. These methods are usually based on local distance measures and are capable of handling large databases [41]:…”

Section: Robust Distance Methodsmentioning

confidence: 99%

“…They proposed the construction of a covariance matrix, based on weighted observations according to their distance from the center. Other robust estimators-such as M-estimator (M stands for maximum likelihood type) and S-estimator (scale estimator), among others-are considered as effective ways to detect multivariate outliers [41,46,47].…”

Section: Robust Distance Methodsmentioning

confidence: 99%

“…Its difficulty lies on the fact that multivariate outlier detection must consider the correlation between the different variables involved in the analysis. This situation may mean some observations that were considered within the acceptable variation range of an individual variable to then become considered as outliers when this correlation is considered [41]. In the process of multivariate outlier detection, researchers may have to deal with two situations.…”

Section: Multivariate Outlier Detectionmentioning

confidence: 99%

See 2 more Smart Citations

Evaluating the Impact of a Two-Stage Multivariate Data Cleansing Approach to Improve to the Performance of Machine Learning Classifiers: A Case Study in Human Activity Recognition

Neira-Rodado

Nugent

Cleland

et al. 2020

Sensors

View full text Add to dashboard Cite

Human activity recognition (HAR) is a popular field of study. The outcomes of the projects in this area have the potential to impact on the quality of life of people with conditions such as dementia. HAR is focused primarily on applying machine learning classifiers on data from low level sensors such as accelerometers. The performance of these classifiers can be improved through an adequate training process. In order to improve the training process, multivariate outlier detection was used in order to improve the quality of data in the training set and, subsequently, performance of the classifier. The impact of the technique was evaluated with KNN and random forest (RF) classifiers. In the case of KNN, the performance of the classifier was improved from 55.9% to 63.59%.

show abstract

Section: Robust Distance Methodsmentioning

confidence: 99%

Section: Robust Distance Methodsmentioning

confidence: 99%

Section: Multivariate Outlier Detectionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluating the Impact of a Two-Stage Multivariate Data Cleansing Approach to Improve to the Performance of Machine Learning Classifiers: A Case Study in Human Activity Recognition

Neira-Rodado

Nugent

Cleland

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Khoshgoftaar et al [12] developed an ensemble filter that includes twenty-five classification techniques, and instances that were incorrectly classified by the majority of classifiers are labeled as noisy. Li et al [15] developed a fuzzy relevance vector machine for learning from imbalanced and noisy data. Zhu and Wu [33] showed that class noise impacts classifiers more severely than attribute noise.…”

Section: Related Workmentioning

confidence: 99%

Class noise detection based on software metrics and ROC curves

Çatal

Alan

Balkan

2011

Information Sciences

View full text Add to dashboard Cite

“…RVM, which was first introduced by Tipping, [17] is often used for classification and pattern recognition [18][19][20]. Previous work concerning classification has been devoted to fault diagnosis of low speed bearing using a multi-class RVM [21].…”

Section: Introductionmentioning

confidence: 99%