Undersampling Method Research Articles

Improvement The Accuracy of Convolutional Neural Network with Using Undersampling Method on Unbalanced Credit Card Dataset

In this study, we address the challenge of imbalanced data in credit card fraud detection by proposing a novel approach that leverages Convolutional Neural Networks (CNNs) and undersampling techniques. The imbalance in the dataset, typical of real-world financial transactions, often leads to biased models favoring the majority class. To mitigate this, we employ undersampling to balance the classes, thereby enhancing the CNN's ability to learn from minority instances crucial for fraud detection. Our method is validated on a large unbalanced credit card dataset, demonstrating significant improvements in accuracy compared to traditional CNN models trained on imbalanced data. We evaluate our approach using standard performance metrics, including precision, recall, and F1-score, showcasing its effectiveness in accurately identifying fraudulent transactions while minimizing false positives. Furthermore, we pro-vide insights into the CNN's decision-making process through visualization techniques, shedding light on its ability to discern fraudulent patterns within the data. Our findings highlight the importance of addressing class imbalance in fraud detection tasks and underscore the efficacy of undersampling in enhancing the performance of deep learning models, particularly CNNs, in handling imbalanced datasets.

Deep adaptive ensemble learning for imbalanced credit card fraud detection

ABSTRACT Traditional techniques for credit card fraud detection often struggle with highly imbalanced datasets, which compounds the complexity of fraudulent transaction detection. To tackle this issue, the current paper puts forth a novel technique known as deep adaptive ensemble learning (DAEL). Initially, we employ the ensemble-based under-sampling method to bring equilibrium to the data. This is followed by the application of a deep autoencoder for the purpose of feature extraction and dimensionality reduction. Then, a bidirectional long short-term memory model is trained using the features extracted from the autoencoder. Finally, an adaptive weighting technique is applied to unify multiple models. The empirical results demonstrate that the proposed DAEL technique significantly surpasses individual models and conventional ensemble learning techniques in performance.

Optimisation of multiple clustering based undersampling using artificial bee colony: Application to improved detection of obfuscated patterns without adversarial training

Attack detection is one of the main features required in modern defence systems. Despite the ongoing research, it remains challenging for a typical mechanism like network-based intrusion detection system (NIDS) to catch up with evolving adversarial attacks. They specifically aim to confuse a machine-learning based predictor. Without the knowledge of adversarial patterns, the best approach is generalising signatures learned from a dataset of legitimate connections and known intrusions. This work focuses on analysing non-payload traffics so that the resulting techniques can be exploited to a range of network-based applications. It investigates a novel means to deal with the problem of imbalanced classes. An optimised undersampling method is introduced to select a subset of majority-class representatives initially created through an ensemble clustering procedure. A weighted combination of criteria representing distributions within and between classes is proposed as the objective function for a global optimisation using the artificial bee colony (ABC). This approach usually outperforms its baselines and other state-of-the-art undersampling models, with ABC being more effective using the global best strategy than a random selection of solutions or an iterative greedy search. The paper also details the parameter analysis offering a heuristic guide for potential taking up of the proposed techniques.

DBOS_US: a density-based graph under-sampling method to handle class imbalance and class overlap issues in software fault prediction

DBOS_US: a density-based graph under-sampling method to handle class imbalance and class overlap issues in software fault prediction

An Undersampling Method Approaching the Ideal Classification Boundary for Imbalance Problems

Data imbalance is a common problem in most practical classification applications of machine learning, and it may lead to classification results that are biased towards the majority class if not dealt with properly. An effective means of solving this problem is undersampling in the borderline area; however, it is difficult to find the area that fits the classification boundary. In this paper, we present a novel undersampling framework, whereby the clustering of samples in the majority class is conducted and segmentation is then performed in the boundary area according to the clusters obtained; this enables a better shape that fits the classification boundary to be obtained via the performance of random sampling in the borderline area of these segments. In addition, we hypothesize that there exists an optimal number of classifiers to be integrated into the method of ensemble learning that utilizes multiple classifiers that have been obtained via sampling to promote the algorithm. After passing the hypothesis test, we apply the improved algorithm to the newly developed method. The experimental results show that the proposed method works well.

Software Defect Prediction Using Deep Q-Learning Network-Based Feature Extraction

Machine learning-based software defect prediction (SDP) approaches have been commonly proposed to help to deliver high-quality software. Unfortunately, all the previous research conducted without effective feature reduction suffers from high-dimensional data, leading to unsatisfactory prediction performance measures. Moreover, without proper feature reduction, the interpretability and generalization ability of machine learning models in SDP may be compromised, hindering their practical utility in diverse software development environments. In this paper, an SDP approach using deep Q-learning network (DQN)-based feature extraction is proposed to eliminate irrelevant, redundant, and noisy features and improve the classification performance. In the data preprocessing phase, the undersampling method of BalanceCascade is applied to divide the original datasets. As the first step of feature extraction, the weight ranking of all the metric elements is calculated according to the expected cross-entropy. Then, the relation matrix is constructed by applying random matrix theory. After that, the reward principle is defined for computing the Q value of Q-learning based on weight ranking, relation matrix, and the number of errors, according to which a convolutional neural network model is trained on datasets until the sequences of metric pairs are generated for all datasets acting as the revised feature set. Various experiments have been conducted on 11 NASA and 11 PROMISE repository datasets. Sensitive analysis experiments show that binary classification algorithms based on SDP approaches using the DQN-based feature extraction outperform those without using it. We also conducted experiments to compare our approach with four state-of-the-art approaches on common datasets, which show that our approach is superior to these methods in precision, F-measure, area under receiver operating characteristics curve, and Matthews correlation coefficient values.

Performance of Random Oversampling, Random Undersampling, and SMOTE-NC Methods in Handling Imbalanced Class in Classification Models

One common challenge in classification modeling is the existence of imbalanced classes within the data. If the analysis continues with imbalanced classes, it is probable that the result will demonstrate inadequate performance when forecasting new data. Various approaches exist to rectify this class imbalance issue, such as random oversampling, random undersampling, and the Synthetic Minority Over-sampling Technique for Nominal and Continuous (SMOTE-NC). Each of these methods encompasses distinct techniques aimed at achieving balanced class distribution within the dataset. Comparison of classification performance on imbalanced classes handled by these three methods has never been carried out in previous research. Therefore, this study undertakes an evaluation of classification models (specifically Gradient Boosting, Random Forest, and Extremely Randomized Trees) in the context of imbalanced class data. The results of this research show that the random undersampling method used to balance the class distribution has the best performance on two classification models (Random Forest and Gradient Boosted Tree).

UsIL-6: An unbalanced learning strategy for identifying IL-6 inducing peptides by undersampling technique

Background and objectiveInterleukin-6 (IL-6) is the critical factor of early warning, monitoring, and prognosis in the inflammatory storm of COVID-19 cases. IL-6 inducing peptides, which can induce cytokine IL-6 production, are very important for the development of diagnosis and immunotherapy. Although the existing methods have some success in predicting IL-6 inducing peptides, there is still room for improvement in the performance of these models in practical application. MethodsIn this study, we proposed UsIL-6, a high-performance bioinformatics tool for identifying IL-6 inducing peptides. First, we extracted five groups of physicochemical properties and sequence structural information from IL-6 inducing peptide sequences, and obtained a 636-dimensional feature vector, we also employed NearMiss3 undersampling method and normalization method StandardScaler to process the data. Then, a 40-dimensional optimal feature vector was obtained by Boruta feature selection method. Finally, we combined this feature vector with extreme randomization tree classifier to build the final model UsIL-6. ResultsThe AUC value of UsIL-6 on the independent test dataset was 0.87, and the BACC value was 0.808, which indicated that UsIL-6 had better performance than the existing methods in IL-6 inducing peptide recognition. ConclusionsThe performance comparison on independent test dataset confirmed that UsIL-6 could achieve the highest performance, best robustness, and most excellent generalization ability. We hope that UsIL-6 will become a valuable method to identify, annotate and characterize new IL-6 inducing peptides.

An effective undersampling method for biomedical named entity recognition using machine learning

An effective undersampling method for biomedical named entity recognition using machine learning

A Novel Spark-Based Attribute Reduction and Neighborhood Classification for Rough Evidence.

Neighborhood classification (NEC) algorithms have been widely used to solve classification problems. Most traditional NEC algorithms employ the majority voting mechanism as the basis for final decision making. However, this mechanism hardly considers the spatial difference and label uncertainty of the neighborhood samples, which may increase the possibility of the misclassification. In addition, the traditional NEC algorithms need to load the entire data into memory at once, which is computationally inefficient when the size of the dataset is large. To address these problems, we propose a novel Spark-based attribute reduction and NEC for rough evidence in this article. Specifically, we first construct a multigranular sample space using the parallel undersampling method. Then, we evaluate the significance of attribute by neighborhood rough evidence decision error rate and remove the redundant attribute on different samples subspaces. Based on this attribute reduction algorithm, we design a parallel attribute reduction algorithm which is able to compute equivalence classes in parallel and parallelize the process of searching for candidate attributes. Finally, we introduce the rough evidence into the classification decision of traditional NEC algorithms and parallelize the classification decision process. Furthermore, the proposed algorithms are conducted in the Spark parallel computing framework. Experimental results on both small and large-scale datasets show that the proposed algorithms outperform the benchmarking algorithms in the classification accuracy and the computational efficiency.

Alleviating Class Imbalance Issue in Software Fault Prediction Using DBSCAN-Based Induced Graph Under-Sampling Method

Alleviating Class Imbalance Issue in Software Fault Prediction Using DBSCAN-Based Induced Graph Under-Sampling Method

A credit scoring ensemble model incorporating fuzzy clustering particle swarm optimization algorithm

The emergence of credit has generated a wealth of data on consumer lending behavior. In recent years, financial institutions have also started to use such data to make informed lending decisions based on fine-grained customer data, but conventional risk assessment models are inadequate in meeting the risk control requirements of the financial industry. Therefore, this paper proposes a credit scoring ensemble model incorporating fuzzy clustering particle swarm optimization (PSO) algorithm to obtain better credit risk prediction capability. First, a weighted outlier detection method based on the Induced Ordered Weighted Average Operator is proposed to preprocess the data to reduce noisy data’s misleading effect on model training. Then, an undersampling method combined with fuzzy clustering PSO is proposed to overcome the negative effect of category imbalance on model training by resampling the data. In addition, a hyperparameter optimization framework is introduced to adaptively adjust important parameters in the ensemble model considering the impact of parameter settings on the training performance of the model. Based on the evaluation metrics of F-score, AUC, and Kappa coefficient, an empirical analysis was conducted on five credit risk datasets. The results show that the proposed method outperforms the comparative model with an improvement of 10% to 50% in terms of F-score and AUC. The highest achieved F-score is 0.9488, and the maximum AUC is 0.9807, demonstrating the effectiveness of the proposed method. The kappa coefficient results indicate a high level of consistency in the predicted classification results of the model.

A Semisupervised Approach for Industrial Anomaly Detection via Self-Adaptive Clustering

With the rapid development of the Industrial Internet of Things (IIoT), log-based anomaly detection has become vital for smart industrial construction that has prompted many researchers to contribute. To detect anomalies based on log data, semi-supervised approaches stand out from supervised and unsupervised approaches because they only require a portion of labeled data and are relatively stable. However, the state-of-the-art semi-supervised approaches still suffer from two main problems: manual parameter setting and unsatisfactory performance with high false positives. We propose AdaLog, an integrated semi-supervised approach based on self-adaptive clustering, for industrial anomaly detection. In particular, the clustering step performs automatic label probability estimation by distinguishing twelve situations so that the label probability of each unlabeled data can be carefully calculated, leading to high accuracy. In addition, AdaLog employs a pre-trained model to learn contextual information comprehensively and a transformer-based model to detect anomalies efficiently. To alleviate class imbalance, an undersampling method is incorporated. The results on three popular datasets demonstrate that AdaLog significantly outperforms three state-of-the-art semi-supervised approaches by 17.8%–2,489.8% on average in terms of F1-score, and is even superior to two supervised approaches in most cases with average improvements of 10.9%–23.8%.

Detecting central apneas using multichannel signals in premature infants.

Objective. Monitoring of apnea of prematurity, performed in neonatal intensive care units by detecting central apneas (CAs) in the respiratory traces, is characterized by a high number of false alarms. A two-step approach consisting of a threshold-based apneic event detection algorithm followed by a machine learning model was recently presented in literature aiming to improve CA detection. However, since this is characterized by high complexity and low precision, we developed a new direct approach that only consists of a detection model based on machine learning directly working with multichannel signals.Approach. The dataset used in this study consisted of 48 h of ECG, chest impedance and peripheral oxygen saturation extracted from 10 premature infants. CAs were labeled by two clinical experts. 47 features were extracted from time series using 30 s moving windows with an overlap of 5 s and evaluated in sets of 4 consecutive moving windows, in a similar way to what was indicated for the two-step approach. An undersampling method was used to reduce imbalance in the training set while aiming at increasing precision. A detection model using logistic regression with elastic net penalty and leave-one-patient-out cross-validation was then tested on the full dataset.Main results. This detection model returned a mean area under the receiver operating characteristic curve value equal to 0.86 and, after the selection of a FPR equal to 0.1 and the use of smoothing, an increased precision (0.50 versus 0.42) at the expense of a decrease in recall (0.70 versus 0.78) compared to the two-step approach around suspected apneic events.Significance. The new direct approach guaranteed correct detections for more than 81% of CAs with lengthL≥ 20 s, which are considered among the most threatening apneic events for premature infants. These results require additional verifications using more extensive datasets but could lead to promising applications in clinical practice.

A majority affiliation based under-sampling method for class imbalance problem

Class imbalance poses difficulties in training a classifier that perform well on minority classes, especially when there is a high imbalance ratio and significant class overlap. Existing data-level methods often suffer from problems like information loss and overfitting. To address these problems, we introduce a novel majority affiliation based under-sampling method (MAUS). The MAUS method employs a support vector data description model to capture the distribution of the minority class, thereby forming a hyper-sphere to establish a majority affiliation for each sample. The high-dimensional hyper-sphere constructed through all minority class samples avoids the problem of overfitting. Leveraging the majority affiliation in conjunction with the k-nearest neighbor algorithm, MAUS is capable of identifying region of class overlap and subsequently removing majority samples within these regions that negatively impact classification performance. This selective removal process minimizes excessive information loss at classification boundaries while alleviating the issue of class overlap. Furthermore, by removing those majority samples that are situated far from the classification boundary, MAUS reduces the imbalance ratio to the expected value, resulting in the attainment of a balanced dataset. To validate the effectiveness of our method, we conducted extensive experiments comparing it with state-of-the-art methods on 30 publicly available datasets. The results indicate that our approach outperforms existing methods on most of datasets and classifiers.

Dietary Inflammatory Index and diabetic retinopathy risk in US adults: findings from NHANES (2005–2008)

BackgroundInflammation is associated with the pathophysiology of diabetic retinopathy (DR). Within the framework of complete dietary patterns, the Dietary Inflammatory Index (DII) was formulated to evaluate the inflammatory properties inherent in a diet. The main purpose of the current study was to assess the relationship between DII and DR using National Health and Nutrition Examination Survey (NHANES).MethodsThe original sample size included 1,148 diabetes patients out of 2005–2008 NHANES surveys. Twenty-four-hour dietary consumptions were used to calculate the DII scores. Demographic characteristics and retina examinations were collected for the comparison between DR and non-DR groups in diabetes patients. The relationship between DII and DR was analyzed by a logistic regression model.Results227 subjects (110 non-DR and 117 DR) were selected in the analyses by using undersampling method to balance the sample size. Compared with non-DR group, DR group had higher DII values (1.14 ± 0.29 vs. 1.49 ± 0.21, p = 0.32), higher levels of HbA1c (6.8 ± 1.1% vs. 7.7 ± 2.6%, p < 0.001), longer duration of diabetes (6.52 ± 12 years vs. 14 ± 11 years, p < 0.001). The odds rate (OR) of DII for DR from the logistic regression was 1.38 (95%CI 1.06–1.81, p < 0.001). HbA1c, diabetes duration and obesity were important influencing factors, and their ORs were 1.81 (95% CI:1.31–2.50), 1.12 (95%CI:1.04–1.20), 4.01 (95%CI:1.12–14.32), respectively. In addition, the most important dietary indices for DR were different across males and females.ConclusionsThe current study demonstrates that a higher DII is associated with an increased risk of DR in US adults. Considering diet as a modifiable factor, limiting pro-inflammatory diets or encouraging an anti-inflammatory diet may be a promising and cost-effective method in the management of DR.

A novel approach for bearings multiclass fault diagnosis fusing multiscale deep convolution and hybrid attention networks

Insufficient and imbalanced samples pose a significant challenge in bearing fault diagnosis, leading to low diagnosis accuracy. However, the fault characteristics of vibration signals are weak and difficult to extract when faults occur in the early stage. This paper proposes an effective fault diagnosis method that addresses small and imbalanced sample problems under noise interference. First, the number of faulty samples in the form of 1D signals is increased mainly by the sliding split sampling method. The preprocessed data are used to create 2D time–frequency diagrams using the continuous wavelet transform (CWT), which can extract effective features to improve the data quality. Subsequently, the minority samples are oversampled by combining synthetic minority oversampling technique to realize time–frequency conversion augmented oversampling. Moreover, the clustering method and random undersampling method are introduced to prevent the overfitting and underfitting problems respectively. Then, we propose a hybrid attention mechanism to enhance the extraction of effective feature information. This combination, integrating CWT with a multicolumn modified deep residual network, effectively extracts fault characteristics and suppresses noise effects. The experimental results demonstrate the effectiveness of the proposed method by comparison with other advanced methods using two case studies of bearing datasets.

Adaptive undersampling and short clip-based two-stream CNN-LSTM model for surgical phase recognition on cholecystectomy videos

Surgical phase recognition is challenging due to overfitting problems caused by imbalanced data among surgical phases. We proposed an adaptive sampling rate-based undersampling method that could generate the number of each surgical phase data similarly to alleviate biased learning. To improve the performance of our method, we also introduced a two-stream CNN-LSTM model that could extract temporal information on behavioral changes between each image frame. First, we extracted a total of 40,236 short clips using an adaptive subsampling rate from the entire video. Each short clip was entered into a pre-trained GoogLeNet. The output with visual information was then immediately fed into a sequence-to-sequence LSTM model to extract temporal information of neighbor frames within a short clip. At the same time, another sequence-to-vector LSTM was used, to extract temporal information from all successive image frames to predict the final surgical phase. The proposed method was evaluated with a public dataset Cholec80. The proposed approach outperformed state-of-the-art methods, showing a high F1-score of 87.12% and an AUC of 98.00%. In addition, the F1-score deviation between all phases decreased by about 10% compared to that before applying undersampling. Experimental results confirmed that employing our proposed method could learn enrich temporal information from short clips. It outperformed the conventional one-stream CNN-LSTM architecture.

A new hybrid credit scoring ensemble model with feature enhancement and soft voting weight optimization

The explosive development of artificial intelligence (AI) has reshaped all aspects of life, including credit scoring. At the same time, the rapid expansion of the consumer finance industry has led to a huge demand. In this study, a new hybrid ensemble model with feature enhancement and soft voting weight optimization is proposed to achieve superior predictive power for credit scoring. For mining and characterizing the implicit information of the features, a new voting-based feature enhancement method is proposed to adaptively integrate the outlier detection and clustering capabilities through the weighted voting mechanism to form a feature-enhanced training set. To balance the feature-enhanced training set precisely and effectively, a new bagging-based undersampling method is proposed to obtain a balanced training set by undersampling from the negative binomial distribution through the bagging strategy. To maximize the performance of the model, a new weight-optimized soft voting method is proposed to optimize the soft voting weights of the base classifiers in the classifier ensemble using the COBYLA algorithm and then constructing the stacking-based ensemble model. Five datasets and five evaluation indicators were used for evaluation. The experimental results demonstrate the superior performance of the proposed model and prove its robustness and effectiveness.

Improving the accuracy of recurrent neural networks models in predicting software bug based on undersampling methods

&lt;span&gt;The process of identifying software bugs is of paramount importance as it ensures software reliability and facilitates maintenance activities. The quality improvement process of software relies heavily on software bug prediction (SBP). In SBP, the task of accurately identifying defective source code poses a significant challenge. Numerous of machine learning (ML) models has been developed specifically to address this challenge in SBP. Nonetheless, the class imbalance issue restricts the potential of these models to predict software bugs accurately. This issue poses a significant hindrance to the efficiency of these models, leading to imbalanced false-positive and false-negative outcomes. Previous studies have paid limited attention to addressing the challenge of class imbalance in SBP. This study aims to fill this research gap by employing a combination of two recurrent neural networks (RNNs), namely &lt;a name="_Hlk141433135"&gt;&lt;/a&gt;long-short-term memory (LSTM) and gated recurrent unit (GRU), along with an undersampling method (near miss) to effectively tackle this issue. Experiments have been conducted on publicly available benchmark datasets, considering both class-level and file-level metrics. The experimental results lead to the conclusion that our models outperform others and the combination of RNNs models with undersampling methods leads to improved bug prediction performance, particularly for datasets with imbalanced class distributions.&lt;/span&gt;