One-class ensemble classifier for data imbalance problems

Hayashi, Toshitaka; Fujita, Hamido

doi:10.1007/s10489-021-02671-1

Cited by 32 publications

(8 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mohammad et al [29] proposed a clustering-based under-sampling method where they used C4.5 bagging and boosting approach for primary classification of pre-processed data, K-Means algorithm to cluster the data, and also Mahalanobis distance is introduced into their proposed method for the purpose of preserving the data distribution pattern in each cluster. Moreover, Hayashi et al [30] proposed a single class ensemble classifier where the model learning separately from majority and minority class, and final classification result is the combination of knowledge from both classes' training.…”

Section: Methods Of Imbalance Classificationmentioning

confidence: 99%

A binary PSO-based ensemble under-sampling model for rebalancing imbalanced training data

Fong

et al. 2021

J Supercomput

View full text Add to dashboard Cite

Ensemble technique and under-sampling technique are both effective tools used for imbalanced dataset classification problems. In this paper, a novel ensemble method combining the advantages of both ensemble learning for biasing classifiers and a new under-sampling method is proposed. The under-sampling method is named Binary PSO instance selection; it gathers with ensemble classifiers to find the most suitable length and combination of the majority class samples to build a new dataset with minority class samples. The proposed method adopts multi-objective strategy, and contribution of this method is a notable improvement of the performances of imbalanced classification, and in the meantime guaranteeing a best integrity possible for the original dataset. We experimented the proposed method and compared its performance of processing imbalanced datasets with several other conventional basic ensemble methods. Experiment is also conducted on these imbalanced datasets using an improved version where ensemble classifiers are wrapped in the Binary PSO instance selection. According to experimental results, our proposed methods outperform single ensemble methods, state-of-the-art under-sampling methods, and also combinations of these methods with the traditional PSO instance selection algorithm.

show abstract

Section: Methods Of Imbalance Classificationmentioning

confidence: 99%

A binary PSO-based ensemble under-sampling model for rebalancing imbalanced training data

Fong

et al. 2021

J Supercomput

View full text Add to dashboard Cite

show abstract

“…Then, testing samples are predicted based on all classifiers in the manner of ensemble learning [20]. Such kind of strategy is called one-class ensemble [21][22][23], which is promising research in detecting unseen samples. Moreover, one-class ensemble is effective for data imbalance problem.…”

Section: One-class Classificationmentioning

confidence: 99%

“…Moreover, one-class ensemble is effective for data imbalance problem. Since the models are trained from one class, data balance is not a problem [23]. Such an approach is applicable in many situations.…”

Section: One-class Classificationmentioning

confidence: 99%

See 1 more Smart Citation

OCFSP: self-supervised one-class classification approach using feature-slide prediction subtask for feature data

Hayashi

Fujita

2022

Soft Comput

Self Cite

View full text Add to dashboard Cite

One-class classification (OCC) is a machine learning problem where training data has only one class. Recently, self-supervised OCC algorithms have been increasing attention. These algorithms train the model for pretext tasks and use the model error for OCC. However, these tasks are specialized for images, and applying them to feature data is not practical or appropriate for such purpose. This paper proposes a one-class classification approach using featureslide prediction (FSP) subtask for feature data (OCFSP). In particular, the selflabeled dataset is created from training data. In which additional feature vectors are generated by sliding original vectors and self-annotated as the number of the feature slide. Such a dataset is used to train a multi-class classifier, which aims to predict the number of the feature slides. Since this classification model is built using data from only one class, the FSP accuracy for seen data is high relative to unseen data. Accordingly, OCC could be made using the accuracy of FSP. Proposed methods are experimented with using the imbalanced-learn dataset.

show abstract

“…For example, it often happens in classification models to have high accuracy but a low recall rate. This is because the trained model does not consider data set imbalances carefully [37], and the ability to manage data imbalances is insufficient. In order to compensate for this shortcoming, different criteria for proving the efficiency of the model from various perspectives are required.…”

Section: Introductionmentioning

confidence: 99%

Email Spam detection in IoT based on Hybrid deep learning model and multi-objective optimization by NSGA II

Dehghani

Ahmadinia

Ghafoori

2023

Preprint

View full text Add to dashboard Cite

The Internet of Things (IoT) connects a range of things, including sensors, physical devices, controllers, and intelligent computer processors. Physical objects with the ability to organize and control independently are referred to as smart devices in the IoT architecture. These smart devices are becoming an integral aspect of human life, from smart homes to large industrial and organizational sectors. Despite the numerous benefits of email regarding information generation and reminders based on predefined regulations, spam emails sent by thingbots pose a potential concern in the Internet of Things. Recently, several studies have used machine learning systems and deep learning models to detect email spam in the internet of things. The presence of unbalanced data, which impacts classification accuracy, is one of the challenges associated with spam detection. In this study, bidirectional gated recurrent unit (BiGRU) and Convolution neural network (CNN) are combined with the Non-dominated Sorting Genetic Algorithm-II (NSGA II) multi-objective optimization method to effectively address imbalance problems. This solution utilizes the two classification criteria TPR and FPR, as NSGA II objective functions and is capable of resolving the imbalance problem in email spam data. The proposed technique is evaluated using the Enron dataset, and the results indicate that the proposed method is more accurate at detecting spam than other baseline methods.

show abstract

One-class ensemble classifier for data imbalance problems

Cited by 32 publications

References 35 publications

A binary PSO-based ensemble under-sampling model for rebalancing imbalanced training data

A binary PSO-based ensemble under-sampling model for rebalancing imbalanced training data

OCFSP: self-supervised one-class classification approach using feature-slide prediction subtask for feature data

Email Spam detection in IoT based on Hybrid deep learning model and multi-objective optimization by NSGA II

Contact Info

Product

Resources

About