Data Imbalance Issue Research Articles

Abstract 4136459: Machine Learning in Time-to-Event Prediction of Ventricular Arrhythmias among Older Adults with Type 2 Diabetes and Coronary Artery Disease

Introduction: Patients with Type 2 diabetes mellitus (T2DM) have an increased risk for coronary artery disease (CAD) compared to patients without T2DM. Ventricular arrhythmias (VA), such as ventricular fibrillation and ventricular tachycardia, are the major causes of mortality among patients with CAD. Thus, T2DM patients with CAD, especially older adults, have a higher risk of VA compared to patients without T2DM. The time-to-event prediction can contribute to decreasing mortality due to VA in older adults with T2DM and CAD by providing an estimated time and probability of diagnosis at specific time points. However, the limited number of older T2DM patients with CAD, who are diagnosed with ventricular arrhythmias, lowers the performance of a traditional time-to-event analysis model. Hypothesis: We hypothesize that machine learning can improve performance in time-to-event prediction. Methods: This study includes 3975 participants aged >65 years, diagnosed with T2DM and CAD in the All of US database. The baseline time was the earliest date when the patients were older than 65 years and diagnosed with T2DM and CAD. Of all participants, 379 were diagnosed with VA within 5 years from baseline. We compared the machine learning–based time-to-event analysis models to the Cox Proportional Hazards model to explore the potential of machine learning in time-to-event analysis. Additionally, the Synthetic Minority Oversampling Technique was applied to resolve the imbalanced data issue. The model performance was evaluated based on the concordance index, which calculates the correlation between predicted risk scores and actual observations, after 5-fold cross-validation. Results: The average concordance index of the Gradient Boosting Machines model was the highest among the five models, including the Cox Proportional Hazards model, with balanced data (Table). Conclusion: We identified that machine learning could improve the performance of prediction models in time-to-event analysis. The Synthetic Minority Oversampling Technique could mitigate the issue of imbalanced data. In addition, machine learning-based survival models showed higher concordance index values compared to the Cox Proportional Hazards model.

Enhancing power transformer health assessment through dimensional reduction and ensemble approaches in Dissolved Gas Analysis

AbstractTransformer health analysis using Dissolved Gas Analysis is crucial for diagnosing power transformer faults. This paper proposes an innovative approach to diagnose power transformer faults by integrating machine learning algorithms with Ensemble techniques. The method involves fusing reduced dimensional input features through Principal Component Analysis with Ensemble techniques such as Bagging, Decorate, and Boosting. Various machine learning algorithms, including Decision Tree (DT), K‐Nearest Neighbours, Radial Basis Function Network, and Support Vector Machine, are employed in conjunction with Ensemble techniques. The long short‐term memory algorithm was used to create synthetic data to solve the issue of data imbalance. A dataset of 683 samples is used in the study for training, testing, validation, and comparison with current techniques. The results highlight the effectiveness of Ensemble techniques, particularly Boosting, which demonstrates superior performance across all classification algorithms. The Boosting with DT algorithm achieves an impressive accuracy of 98.32%, surpassing alternative methods. In validation, the proposed Boosting Ensemble technique outperforms various approaches, showcasing its diagnostic accuracy and superiority over alternative methods. The research emphasises the model's effectiveness in smoothing input vectors, enhancing harmony with ensemble techniques, and overcoming limitations in prior methods.

Improving Sentiment Analysis Accuracy Using CRNN on Imbalanced Data: a Case Study of Indonesian National Football Coach

Conducting sentiment research on the perception of the Indonesian people towards Shin Tae Yong's (STY) role as coach of the Indonesian National Football Team (PSSI) is crucial as it can assist PSSI in determining whether to extend STY's contract. Prior studies have demonstrated that Deep Learning achieves a high level of accuracy when applied to sentiment analysis in many domains. Nevertheless, no investigation has been conducted thus far utilizing deep learning techniques to examine emotion towards STY. This study employs modified Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Convolutional Recurrent Neural Networks (CRNN), and CRNN models with and without data oversampling. The research findings indicate that the CRNN model, when combined with data oversampling and a redesigned architecture, achieves the highest level of accuracy (1.00) and consistently performs well. This research provides significant contributions in three areas: firstly, it utilizes Deep Learning techniques for sentiment analysis on STY; secondly, it modifies the CRNN architecture; and thirdly, it applies data oversampling to address the issue of imbalanced data.

Securing transactions: a hybrid dependable ensemble machine learning model using IHT-LR and grid search

Financial institutions and businesses face an ongoing challenge from fraudulent transactions, prompting the need for effective detection methods. Detecting credit card fraud is crucial for identifying and preventing unauthorized transactions. While credit card fraud incidents are relatively rare, they can result in substantial financial losses, particularly due to the high monetary value associated with fraudulent transactions. Timely detection of fraud enables investigators to take swift actions to mitigate further losses. However, the investigation process is often time-consuming, limiting the number of alerts that can be thoroughly examined each day. Therefore, the primary objective of a fraud detection model is to provide accurate alerts while minimizing false alarms and missed fraud cases. In this paper, we introduce a state-of-the-art hybrid ensemble (ENS) dependable machine learning (ML) model that intelligently combines multiple algorithms with proper weighted optimization using grid search, including decision tree (DT), random forest (RF), K-nearest neighbor (KNN), and multilayer perceptron (MLP), to enhance fraud identification. To address the data imbalance issue, we employ the instant hardness threshold (IHT) technique in conjunction with logistic regression (LR), surpassing conventional approaches. Our experiments are conducted on a publicly available credit card dataset comprising 284,807 transactions. The proposed model achieves impressive accuracy rates of 99.66%, 99.73%, 98.56%, and 99.79%, and a perfect 100% for the DT, RF, KNN, MLP and ENS models, respectively. The hybrid ensemble model outperforms existing works, establishing a new benchmark for detecting fraudulent transactions in high-frequency scenarios. The results highlight the effectiveness and reliability of our approach, demonstrating superior performance metrics and showcasing its exceptional potential for real-world fraud detection applications.

DyConfidMatch: Dynamic thresholding and re-sampling for 3D semi-supervised learning

Semi-supervised learning (SSL) leverages limited labeled and abundant unlabeled data but often faces challenges with data imbalance, especially in 3D contexts. This study investigates class-level confidence as an indicator of learning status in 3D SSL, proposing a novel method that utilizes dynamic thresholding to better use unlabeled data, particularly from underrepresented classes. A re-sampling strategy is also introduced to mitigate bias towards well-represented classes, ensuring equitable class representation. Through extensive experiments in 3D SSL, our method surpasses state-of-the-art counterparts in classification and detection tasks, highlighting its effectiveness in tackling data imbalance. This approach presents a significant advancement in SSL for 3D datasets, providing a robust solution for data imbalance issues.

LuminanceGAN: Controlling the brightness of generated images for various night conditions

There are diverse datasets available for training deep learning models utilized in autonomous driving. However, most of these datasets are composed of images obtained in day conditions, leading to a data imbalance issue when dealing with night condition images. Several day-to-night image translation models have been proposed to resolve the insufficiency of the night condition dataset, but these models often generate artifacts and cannot control the brightness of the generated image. In this study, we propose a LuminanceGAN, for controlling the brightness degree in night conditions to generate realistic night image outputs. The proposed novel Y-control loss converges the brightness degree of the output image to a specific luminance value. Furthermore, the implementation of the self-attention module effectively reduces artifacts in the generated images. Consequently, in qualitative comparisons, our model demonstrates superior performance in day-to-night image translation. Additionally, a quantitative evaluation was conducted using lane detection models, showing that our proposed method improves performance in night lane detection tasks. Moreover, the quality of the generated indoor dark images was assessed using an evaluation metric. It can be proven that our model generates images most similar to real dark images compared to other image translation models.

Deep learning-powered efficient characterization and quantification of microplastics

Characterizing and quantifying microplastics (MPs) are time-consuming and labor-intensive tasks traditionally. This paper presents an artificial intelligence (AI) framework aiming to automate these tasks by integrating computer vision and deep learning techniques. The approach leverages Fourier Transform Infrared (FTIR) spectra and visual images. Primary novelties of this research involve the development of: (1) an AI framework integrating efforts of data processing, analytics, visualization, and human-computer interaction; (2) a method for transforming FTIR data into contour images; (3) data augmentation strategies for resolving data scarcity and imbalance issues; (4) deep learning models for identifying MPs; (5) computer vision algorithms for quantifying MPs; and (6) an engineer-friendly graphic user interface (GUI) for enhancing data accessibility. The AI framework has been applied to polyethylene, polypropylene, polystyrene, polyamide, ethylene-vinyl acetate, and cellulose acetate. Results confirmed the efficacy of the framework, exhibiting high accuracy scores in classification (98 %), segmentation (99 %), and quantification (96 %) tasks. This research advances the capability of automatic assessment of MPs.

Enhancing SQL Injection Attack Detection Using Naïve Bayes and SMOTE Method on Imbalanced Datasets

SQL injection attack detection is a crucial aspect of cybersecurity, considering the potential damage that such attacks can cause. This study aims to evaluate the effectiveness of the Naive Bayes model in detecting SQL injection attacks on an imbalanced dataset. To address the data imbalance issue, the SMOTE (Synthetic Minority Over-sampling Technique) method was applied. The study consists of two phases: first, training and testing the Naive Bayes model on the original dataset without SMOTE, and second, training and testing on the dataset with SMOTE applied. The results indicate that the Naive Bayes model on the dataset without SMOTE achieved an accuracy of 0.9948, F1 Score of 0.9885, Precision of 0.9906, and Recall of 0.9946. After applying SMOTE, the model's performance improved significantly, with an accuracy of 0.9950, F1 Score of 0.9950, Precision of 0.9950, and Recall of 0.9950. This improvement suggests that SMOTE effectively enhanced class balance in the dataset, improving the model's ability to detect both malicious and safe queries. The study recommends exploring other resampling methods, feature engineering analysis, and testing on more diverse datasets as well as implementation in real-world environments for future research.

Crispr-SGRU: Prediction of CRISPR/Cas9 Off-Target Activities with Mismatches and Indels Using Stacked BiGRU

CRISPR/Cas9 is a popular genome editing technology, yet its clinical application is hindered by off-target effects. Many deep learning-based methods are available for off-target prediction. However, few can predict off-target activities with insertions or deletions (indels) between single guide RNA and DNA sequence pairs. Additionally, the analysis of off-target data is challenged due to a data imbalance issue. Moreover, the prediction accuracy and interpretability remain to be improved. Here, we introduce a deep learning-based framework, named Crispr-SGRU, to predict off-target activities with mismatches and indels. This model is based on Inception and stacked BiGRU. It adopts a dice loss function to solve the inherent imbalance issue. Experimental results show our model outperforms existing methods for off-target prediction in terms of accuracy and robustness. Finally, we study the interpretability of this model through Deep SHAP and teacher–student-based knowledge distillation, and find it can provide meaningful explanations for sequence patterns regarding off-target activity.

Unravelling sleep patterns: Supervised contrastive learning with self-attention for sleep stage classification

Sleep data scoring is a crucial and primary step for diagnosing sleep disorders to know the sleep stages from the PSG signals. This study uses supervised contrastive learning with a self-attention mechanism to classify sleep stages. We propose a deep learning framework for automatic sleep stage classification, which involves two training phases: (1) the feature representation learning phase, in which the feature representation network (encoder) learns to extract features from the electroencephalogram (EEG) signals, and (2) the classification network training phase, where a pre-trained encoder (trained during phase I) along with the classifier head is fine-tuned for the classification task. The PSG data shows a non-uniform distribution of sleep stages, with wake (W) (around 30% of total samples) and N2 stages (around 58% and 37% of total samples in Physionet EDF-Sleep 2013 and 2018 datasets, respectively) being more prevalent, leading to an imbalanced dataset. The imbalanced data issue is addressed using a weighted softmax cross-entropy loss function that assigns higher weights to minority sleep stages. Additionally, an oversampling technique (the synthetic minority oversampling technique (SMOTE) (Chawla et al., 2002)[1] ) is applied to generate synthetic samples for minority classes. The proposed model is evaluated on the Physionet EDF-Sleep 2013 and 2018 datasets using Fpz-Cz and Pz-Oz EEG channels. It achieved an overall accuracy of 94.1%, a macro F1 score of 92.64, and a Cohen’s Kappa coefficient of 0.92. Ablation studies demonstrated the importance of triplet loss-based pre-training and oversampling for enhancing performance. The proposed model requires minimal pre-processing, eliminating the need for extensive signal processing expertise, and thus is well-suited for clinicians diagnosing sleep disorders.

A novel label-aware global graph construction method and spiking-coded graph neural network for intelligent process fault diagnosis

Fault diagnosis plays a crucial role in ensuring the safety and efficiency of industrial processes. However, traditional techniques often face difficulties in handling large-scale data characterized by complex structures and relationships. To efficiently represent industrial data as graphs and develop a low-energy-cost feature extraction model, a novel label-aware global graph construction method and a spiking graph convolutional network (SGCN) are proposed in this study to achieve intelligent process fault diagnosis. The label-aware method enhances graph data representation by capturing intrinsic correlations and global features. The SGCN integrates graph convolutional layers with spiking encoding, enabling effective feature extraction while offering computational efficiency advantages. A weighted loss function is introduced to mitigate data imbalance issues. Experiments on the Tennessee Eastman process, the Three-phase Flow Facility, and the de-propanizer distillation process demonstrate SGCN’s superior performance over baseline models in various fault scenarios, while significantly reducing computational costs. The proposed method offers promising potential for reliable and efficient fault diagnosis in complex real-world industrial environments.

MS-CGAN: Fusion of conditional generative adversarial networks and multi-scale spatio-temporal features for lithology identification

Lithology identification constitutes a crucial undertaking in formation evaluation and reservoir characterization. However, the need for improved precision arises in conventional lithology identification models due to the difficulties presented by unequal distributions of small-sample logging data. An effective combination of domain expertise and data-driven models to predict lithology is essential due to the intricate and nonlinear connection between logging parameters and lithology, combined with the distinct characteristics of the oilfield environments. In this paper, we proposed a multi-scale conditional generative adversarial network(MS-CGAN) method, which combines conditional generative adversarial networks with multi-scale spatio-temporal features to address data imbalance issues and enhance the accuracy of lithology classification. Our approach, tested on two small datasets from the Hugoton and Panoma fields, USA, and the Daqing production wells, China, stands out as the optimal choice compared to other models. Comprehensive evaluation results indicate promising practical applications and potential benefits of the new model in enhancing lithology identification using limited data.

LungNeXt: A novel lightweight network utilizing enhanced mel-spectrogram for lung sound classification

Lung auscultation is essential for early lung condition detection. Categorizing adventitious lung sounds requires expert discrimination by medical specialists. This paper details the features of LungNeXt, a novel classification model specifically designed for lung sound analysis. Furthermore, we propose two auxiliary methods: RandClipMix (RCM) for data augmentation and Enhanced Mel-Spectrogram for Feature Extraction (EMFE). RCM addresses the issue of data imbalance by randomly mixing clips within the same category to create new adventitious lung sounds. EMFE augments specific frequency bands in spectrograms to highlight adventitious features. These contributions enable LungNeXt to achieve outstanding performance. LungNeXt optimally integrates an appropriate number of NeXtblocks, ensuring superior performance and a lightweight model architecture. The proposed RCM and EMFE methods, along with the LungNeXt classification network, have been evaluated on the SPRSound dataset. Experimental results revealed a commendable score of 0.5699 for the lung sound five-category task on SPRSound. Specifically, the LungNeXt model is characterized by its efficiency, with only 3.804M parameters and a computational complexity of 0.659G FLOPS. This lightweight and efficient model is particularly well-suited for applications in electronic stethoscope back-end processing equipment, providing efficient diagnostic advice to physicians and patients.

A precise machine learning model: Detecting cervical cancer using feature selection and explainable AI

Cervical cancer is a cancer that remains a significant global health challenge all over the world. Due to improper screening in the early stages, and healthcare disparities, a large number of women are suffering from this disease, and the mortality rate increases day by day. Hence, in these studies, we presented a precise approach utilizing six different machine learning models (decision tree, logistic regression, naïve bayes, random forest, k nearest neighbors, support vector machine), which can predict the early stage of cervical cancer by analysing 36 risk factor attributes of 858 individuals. In addition, two data balancing techniques—Synthetic Minority Oversampling Technique and Adaptive Synthetic Sampling—were used to mitigate the data imbalance issues. Furthermore, Chi-square and Least Absolute Shrinkage and Selection Operator are two distinct feature selection processes that have been applied to evaluate the feature rank, which are mostly correlated to identify the particular disease, and also integrate an explainable artificial intelligence technique, namely Shapley Additive Explanations, for clarifying the model outcome. The applied machine learning model outcome is evaluated by performance evaluation matrices, namely accuracy, sensitivity, specificity, precision, f1-score, false-positive rate and false-negative rate, and area under the Receiver operating characteristic curve score. The decision tree outperformed in Chi-square feature selection with outstanding accuracy with 97.60%, 98.73% sensitivity, 80% specificity, and 98.73% precision, respectively. During the data imbalance, DT performed 97% accuracy, 99.35% sensitivity, 69.23% specificity, and 97.45% precision. This research is focused on developing diagnostic frameworks with automated tools to improve the detection and management of cervical cancer, as well as on helping healthcare professionals deliver more efficient and personalized care to their patients.

CRATSM: An Effective Hybridization of Deep Neural Models for Customer Retention Prediction in the Telecom Industry

In the dynamic field of Customer Retention Prediction (CRP), strategic marketing and promotion efforts targeting specific customers are crucial. Understanding customer behavior and identifying churn indicators are vital for devising effective retention strategies. However, identifying customers likely to terminate services presents a challenge, leading to data imbalance issues. Existing CRP studies using Machine Learning (ML) techniques and data imbalance methods face problems such as overfitting and computational complexity. Similarly, recent CRP studies employing Deep Learning (DL) approaches rely on data sampling techniques, which can result in overfitting and a lack of cost sensitivity. Additionally, DL approaches struggle with slow convergence and get stuck in local minima. This paper introduces an effective hybrid of Deep Learning (DL) classifiers focusing on cost-metric integration to address data imbalance issues and period-shift Cosine Annealing Learning Rate (ps-CALR) to accelerate model training, ultimately enhancing performance. Three Telecom datasets, namely IBM, Iranian, and Orange, were used to assess the model performance. Empirical findings show that the hybrid DL classifiers significantly improved CRP over conventional ML. This paper contributes methodological advancements and practical insights for effective customer retention in the telecom industry.

Compact convolutional transformers- generative adversarial network for compound fault diagnosis of industrial robot

The safe operation of Industrial robots is a major concern in intelligent manufacturing. Accurate compound fault diagnosis is essential to the safe operation of industrial robots, while it is challenging to achieve since the compound fault samples are hard to be collected. Generative adversarial network (GAN) is a useful tool for addressing the data imbalance issue. However, the computation efficiency of GAN in addressing the data imbalance issue has not been investigated. Hence, this study proposes a lightweight GAN named compact convolutional Transformers-GAN (CCT-GAN) to alleviate the data imbalance issue in compound fault diagnosis modelling. Firstly, the feedback current signals collected from the industrial robot are transformed into time-frequency images via continuous wavelet transformation (CWT). Secondly, CCT-GAN is designed to achieve high-quality fake data generation and compound fault diagnosis modelling without large computational costs. Thirdly, the relation between a single fault and the compound fault is considered in the compound fault diagnosis modelling via multi-hot representation to alleviate the data imbalance issue. An experimental study based on the real-world compound fault dataset of industrial robots reveals that the proposed CCT-GAN shows merits in compound fault diagnosis modelling in comparison with the prevailing algorithms. The results indicate that CCT-GAN can performance of compound fault diagnosis when only 100 data samples from each compound fault category are available.

Improved phase prediction of high-entropy alloys assisted by imbalance learning

Predicting phase formation is crucial in novel high-entropy alloys (HEAs) design. Herein, machine learning and imbalance learning algorithms were combined together to improve the phase prediction of HEAs. In this work, an extensive database by collecting experimental data from published literature was constructed, and the key features affecting the phase formation of HEAs were filtered out by performing a three-step feature selection process. Then, extreme gradient boosting (XGB) models were constructed to categorize phase structures of HEAs with high accuracies. Moreover, the Synthetic Minority Oversampling TEchnique (SMOTE) algorithm was employed for data oversampling to address the data imbalance issue. It was found that imbalanced learning significantly improves the phase prediction, particularly for the minority class, without costing the overall prediction accuracy. Finally, a machine learning-base protocol was proposed to integrate established models to classify the phase formation of HEAs into seven phase labels, and its generalization ability was verified. The present work provides a practical approach in predicting phase structures of HEAs and enhances the efficiency in developing advanced HEAs.

Leveraging machine learning for predictive analysis of tuberculosis treatment outcomes: A comprehensive study using Karnataka TB data

Tuberculosis (TB) remains a global health threat, ranking among the leading causes of mortality worldwide. This study aims to investigate the application of several machine learning models in predicting tuberculosis (TB) treatment outcomes. The study utilizes an extensive dataset from Karnataka, India. The study utilized advanced methodologies, such as XGBoost, to address the issue of imbalanced data and enhance the precision of the predictions. The findings show that these models can greatly improve the identification of patients who are likely to have poor adherence to their treatments, allowing for more focused interventions. The objective is to predict tuberculosis treatment outcomes and assist public health initiatives in Karnataka through the integration of these predictive technologies into the healthcare system.

Phonocardiogram Signal Analysis Based Premature Cardiac Influence Detection using Resnet50 CNN for Public Health Protection

Phonocardiogram (PCG) signal analysis plays a crucial role in the early detection of cardiac abnormalities, which is essential for public health protection. The premature disease prediction is tedious because the feature analysis takes place dimensionality problems which leads poor accuracy. The prevailing ensemble learning models’ methods doesn’t concentrate disease margin facts and properties during classification produce higher false rate because for adjusting feature margins gets lower accuracy. To resolve this problem, we propose a efficient Fuzzy c-means feature modality based Resnet50-RNN to detect premature cardiac influence detection for improving public health. The research proposes a method for premature cardiac influence detection using a ResNet50 Convolutional Neural Network (CNN) model for early prediction of heart disease. To address the issue of imbalanced data, Synthetic Minority Oversampling Technique (SMOTE) is employed for preprocessing. The Cardiac Disease Influence Rate (CDIR) is then used to identify the maximum deficiency-affected feature margins. Fuzzy c-means is utilized to select the mutual features associated with the heart disease influence rate. The ResNet50 CNN model is employed for classification, enabling the prediction of disease margins by risk category. The proposed system effectively identifies the disease based on the disease-affected scaling margin, leading to early disease impact prediction with higher accuracy, specificity, F1-measure, and lower false rates compared to other existing systems. The implementation also demonstrates reduced time complexity.

Intra-pulse modulation discrimination using a self-supervised attention-driven CNN-BiLSTM-VAE combination

Identification of intra-pulse modulation (IPM) of radar signals is a crucial part of contemporary electronic support systems and electronic intelligence reconnaissance. Artificial intelligence (AI)-based methods can be very effective in recognising the IPM of radar signals. In this direction, an automatic method is proposed for recognising a few IPMs of radar signals based on continuous wavelet transform (CWT) and a hybrid model of self-attention (SA)-aided convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM). Firstly, time–frequency attributes of different radar signals are obtained using CWT, and thereafter CNN-SA-BiLSTM is utilised for feature extraction from the 2D scalograms formed by the time–frequency components. The CNN extracts features from the scalograms, SA enhances the discriminative power of the feature map, and BiLSTM detects radar signals based on these features. Additionally, the study addresses real-world data imbalance issues by incorporating a generative AI model, namely the Variational Autoencoder (VAE). The VAE-based approach effectively mitigates challenges arising from data imbalance situations. This method is tested at varying noise levels to give a proper representation of the actual electronic warfare environment. The simulation results demonstrate that the best overall recognition accuracy of the proposed method is 98.4%, even at low signal-to-noise ratios (SNR).