Related references
Note: Only part of the references are listed.
Article
Computer Science, Artificial Intelligence
Karim El Moutaouakil et al.
Summary: This paper proposes Optimal Entropy Genetic Fuzzy-C-Means SMOTE (OEGFCM-SMOTE), a method based on the original mathematical model, soft clustering, and evolutionary optimization, for handling the classification problems of unbalanced data sets. OEGFCM-SMOTE addresses the sensitivity issue of the Kmeans method by generating synthetic samples in safe regions based on Fuzzy-C-Means, and it selects optimal parameters based on entropy to minimize the noise. Experimental results demonstrate that OEGFCM-SMOTE consistently outperforms other popular oversampling methods in multiple performance measures.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Jinjun Ren et al.
Summary: Class-imbalanced classification is a challenging problem where traditional classifiers exhibit bias towards majority classes and generate incorrect predictions. Existing algorithms struggle with the issue of class overlapping. This paper proposes a grouping scheme for minority class samples based on their possibilities of appearing in overlapping regions in the feature space. A new oversampling method is then proposed to generate samples far away from the overlapping region and rectify the decision boundary. An effective classification algorithm for imbalanced data is developed based on these techniques. Extensive experiments demonstrate the superiority of the proposed algorithm over seventeen benchmark algorithms, particularly on highly imbalanced datasets.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Reshma Rastogi et al.
Summary: This article proposes a method for addressing the class imbalance problem in multi-label learning with missing labels. The method constructs a label weight matrix and utilizes discriminatory label weights and auxiliary label correlations to guide the completion of missing labels and learning of the multi-label classifier.
NEURAL PROCESSING LETTERS
(2023)
Article
Medicine, General & Internal
Ricky Walsh et al.
Summary: Tools based on deep learning models have been developed to assist radiologists in diagnosing breast cancer from mammograms. However, the imbalance of malignant and benign samples in the training datasets can lead to biased models. This study evaluates different techniques to address this class imbalance issue and shows that they can counteract the bias towards the majority class. However, these techniques do not improve the model's performance in terms of AUC-ROC, except for the synthetic lesion generation approach.
Article
Computer Science, Interdisciplinary Applications
Shuangyuan Shi et al.
Summary: This paper proposes a no-tardiness job shop scheduling problem with overtime work consideration (NTJSSP-OW) to minimize the total earliness inventory and overtime work costs simultaneously. A mathematical model is formulated and a hybrid genetic algorithm with simulated annealing (GASA) is proposed to solve it. Comprehensive experiments and statistical tests show that GASA algorithm has faster convergence and better global search ability compared to other competing algorithms.
COMPUTERS & INDUSTRIAL ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Cuiqing Jiang et al.
Summary: This paper discusses the solutions for the class imbalance problem in credit scoring, compares the performance of traditional approaches and generative adversarial networks (GANs) in credit scoring, and provides some recommendations with the help of benchmark analysis.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Mi Yan et al.
Summary: This study proposes a deep metric learning-based pseudo-labeling (DML-PL) framework that addresses both class imbalance and insufficient labeled data problems. Through an iterative self-training strategy, a deep metric network is trained to learn compact feature representations of labeled and unlabeled data, generating reliable pseudo-labels and improving training accuracy.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Xiaohan Yuan et al.
Summary: In this paper, a novel and simple convex hull-based SMOTE (CHSMOTE) algorithm is proposed to overcome the weaknesses of SMOTE and alleviate class imbalance problem. CHSMOTE selects the border minority samples as initial samples, identifies the synthesis area based on convex hull, and generates more effective samples by enlarging the generation range. Extensive experiments demonstrate the effectiveness and superiority of the proposed algorithm.
INFORMATION SCIENCES
(2023)
Review
Computer Science, Information Systems
Abiodun M. Ikotun et al.
Summary: Advances in data collection techniques have enabled the accumulation of large quantities of data. The K-means algorithm, while popular, has challenges such as determining the number of clusters and detecting non-Euclidean shapes. Research efforts have been made to improve its performance and robustness.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Tianjun Li et al.
Summary: This paper proposes a new over-sampling method called Subspace-based Minority Over-Sampling (SMO) to address the class imbalance problem. The SMO approach extracts common and unique characteristics of each category of samples using subspace and achieves balanced data by over-sampling the common part and expanding the unique part. Experimental results show that SMO outperforms classical and newly designed over-sampling algorithms and can be used to generate simple images.
INFORMATION SCIENCES
(2023)
Article
Chemistry, Multidisciplinary
Amal Al Ali et al.
Summary: This study aims to develop a better Financial Statement Fraud (FSF) detection model using data from publicly available financial statements of firms in the MENA region. An FSF model was developed using the XGBoost algorithm, which outperformed other algorithms in this study. The class imbalance in the dataset was addressed using the SMOTE algorithm. The optimized XGBoost algorithm achieved a final accuracy of 96.05% in detecting FSF.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Artificial Intelligence
Biao Zhang et al.
Summary: The article proposes a SMOTE-RFE-XGBoost model, which uses the physical angle of human bone as a research index to predict spinal diseases. The model utilizes the SMOTE algorithm to handle category imbalance, and employs LASSO, tree-based feature selection, and recursive feature elimination for feature selection. Various classification models are used to classify the samples and rank the feature importance. The SMOTE-RFE-XGBoost model achieves the best classification performance with an accuracy of 97.56%, MSE of 0.1111, and F1 value of 0.8696. The indicators of lumbar slippage, cervical tilt, pelvic radius, and pelvic tilt are found to be more important.
PEERJ COMPUTER SCIENCE
(2023)
Article
Physics, Multidisciplinary
Cui Fu et al.
Summary: In this paper, a novel relative density-based intuitionistic fuzzy support vector machine (RIFSVM) algorithm is proposed for imbalanced learning in the presence of noise and outliers. The algorithm estimates the intuitionistic fuzzy numbers using the relative density and assigns different fuzzy values to majority class and minority class instances to improve classification performance.
Article
Computer Science, Artificial Intelligence
Ashhadul Islam et al.
Summary: This study introduces an advanced algorithm called KNNOR to address class imbalance by studying the compactness and location of the minority class, identifying critical and safe areas for augmentation, and generating synthetic data points. Experimental results show that the proposed method consistently outperforms other state-of-the-art oversamplers on several common imbalanced datasets, making it easy to use and open source as a python library.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Yuxi Xie et al.
Summary: In this study, a new data resampling technique called Gaussian Distribution based Oversampling (GDO) is proposed to handle imbalanced data for classification. Experimental results show that GDO outperforms other compared methods in terms of AUC, G-mean, and memory usage, with an increase in running time. The effectiveness of GDO is further validated in two real imbalanced data classification problems.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Information Systems
Aimin Zhang et al.
Summary: In recent years, class imbalance learning has gained importance in machine learning. The Synthetic Minority Oversampling TEchnique (SMOTE) is a popular algorithm for addressing class imbalance, but it suffers from noise propagation. To overcome this, researchers have proposed various SMOTE variants. This paper introduces a robust and universal SMOTE hybrid variant algorithm called SMOTE-RkNN, which identifies noise using probability density instead of local neighborhood information.
INFORMATION SCIENCES
(2022)
Article
Chemistry, Analytical
Fayez Alharbi et al.
Summary: Human activity recognition using wearable sensors is a popular research topic in machine learning, facing challenges such as class imbalance. Three hybrid sampling strategies are introduced to generate diverse synthetic samples and improve classification performance. The proposed methods show significant enhancement compared to baseline and constituent techniques in handling class imbalance.
Article
Automation & Control Systems
Shixiang Lu et al.
Summary: This study proposes a class-imbalanced privacy-preserving federated learning framework for fault diagnosis of decentralized wind turbines. By utilizing biometric authentication and privacy-enhancing techniques, it achieves high potential privacy and security. The framework also integrates a gradient-based self-monitoring scheme to enhance the understanding of global information for class-imbalanced fault diagnosis.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Information Systems
Xiangrui Chao et al.
Summary: Balancing the accuracy rates of majority and minority classes is challenging in imbalanced classification. This study introduces a new criterion, an efficiency curve, to comprehensively evaluate imbalanced classifiers and analyzes the impact of imbalanced ratio and data characteristics on classifier efficiency.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Aimin Zhang et al.
Summary: This study presents the instance weighted SMOTE (IW-SMOTE) algorithm, which improves the SMOTE algorithm by indirectly exploiting distribution data. It uses an UnderBagging-like undersampling ensemble algorithm to classify each training instance and acquire confusing information. Based on the confusing information, the algorithm accurately estimates the location information of each instance and handles noisy and borderline instances accordingly. The balanced instance set is then used to train multiple classifiers to verify the algorithm's generality and effectiveness.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Wuxing Chen et al.
Summary: Broad Learning System (BLS) is a fast learning neural network that has shown good performance in various applications. However, conventional BLS has limitations in dealing with class imbalance and parameter tuning. To overcome these challenges, we propose a double-kernelized weighted broad learning system (DKWBLS) that improves feature representation and addresses class imbalance. Experimental results demonstrate the superiority of DKWBLS in handling imbalanced data.
NEURAL COMPUTING & APPLICATIONS
(2022)
Review
Computer Science, Information Systems
Sudhansu R. Lenka et al.
Summary: Credit scoring analysis is of great importance to researchers and the financial industry, as it helps identify deserving applicants for loans with minimal risks. Developing an accurate credit scoring model is challenging due to class imbalance and irrelevant features. Recent research has shown that ensemble learning is superior in this field. This paper conducts a comprehensive comparative analysis of ensemble algorithms to improve oversampling and feature selection techniques. Three feature selection techniques, information gain, principal component analysis, and genetic algorithm, are used to identify relevant features. The experimental results demonstrate that the GA-based FS technique and CatBoost algorithm outperform other models in terms of accuracy, area under the curve, F1-score, Brier score, and Kolmogorov-Smirnov.
WIRELESS COMMUNICATIONS & MOBILE COMPUTING
(2022)
Article
Automation & Control Systems
Junwei Ma et al.
Summary: This study proposes a systematic framework combining k-fold cross-validation, metaheuristics, support vector regression, and statistical tests to improve the reliability and performance of geohazard modeling. By comparing different algorithms, the multiverse optimizer is identified as one of the best-performing algorithms.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Article
Computer Science, Information Systems
Shengnan Shi et al.
INFORMATION SCIENCES
(2022)
Article
Energy & Fuels
Shiqian Wang et al.
Summary: This article presents an improved selective ensemble learning approach to handle class imbalance and base classifier redundancy in load classification. Experimental results show that the approach is effective for load classification tasks.
FRONTIERS IN ENERGY RESEARCH
(2022)
Article
Computer Science, Information Systems
Ahmed Arafa et al.
Summary: Machine learning classifiers often struggle with imbalanced datasets, which are common in real-world scenarios. This paper presents RN-SMOTE, a preprocessing technique that addresses imbalanced classification by oversampling the minority class, removing noise, and rebalancing the dataset. Experimental results demonstrate that RN-SMOTE significantly improves classifier performance compared to the original data and traditional oversampling techniques.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Theory & Methods
Rick Sauber-Cole et al.
Summary: The existence of class imbalance in a dataset can bias the classifier towards majority classification. Generative Adversarial Networks (GANs) have been used to generate instances of the underrepresented class(es) to mitigate this issue. While most research focuses on their application in computer vision tasks, GANs are also being used for tabular data with traditional structured data types.
JOURNAL OF BIG DATA
(2022)
Article
Computer Science, Information Systems
Baiyun Chen et al.
Summary: Imbalanced classification is an important task in supervised learning, and the proposed self-adaptive robust SMOTE, RSMOTE, addresses this issue effectively by introducing relative density to adaptively divide minority samples and reweigh the number of samples needed to be generated based on their chaotic level. Experimental results show that RSMOTE outperforms comparison methods across various metrics, indicating its superiority in handling imbalanced classification with label noise.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Ling Li et al.
Summary: This paper addresses the small dataset problem by developing an information expansion function and a Gaussian mixture model based virtual sample generation method, which improves modeling performance.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Saptarshi Bej et al.
Summary: This article introduces a method, LoRAS, that overcomes the limitations of SMOTE oversampling technique, and through experiments, proves that LoRAS generates better machine learning models on imbalanced datasets, improving F1-Score and balanced accuracy. Compared to most SMOTE extensions, LoRAS achieves better results in generating classification models.
Article
Computer Science, Information Systems
Intouch Kunakorntum et al.
Article
Computer Science, Information Systems
Georgios Douzas et al.
INFORMATION SCIENCES
(2019)
Article
Computer Science, Information Systems
Ke Cheng et al.
Article
Computer Science, Information Systems
Georgios Douzas et al.
INFORMATION SCIENCES
(2018)
Article
Computer Science, Artificial Intelligence
Isaac Triguero et al.
International Journal of Computational Intelligence Systems
(2017)
Article
Computer Science, Artificial Intelligence
Jair Cervantes et al.
Article
Computer Science, Information Systems
Jose A. Saez et al.
INFORMATION SCIENCES
(2015)
Article
Computer Science, Artificial Intelligence
M. A. H. Farquad et al.
DECISION SUPPORT SYSTEMS
(2012)