Handle Data Imbalance Research Articles

Semi-Supervised Intrusion Detection Based on Stacking and Feature-Engineering to Handle Data Imbalance

Objectives: To design an architecture that can effectively handle the imbalance levels and complexities in the network data to provide qualitative predictions. Methods: Experiments were performed with KDD CUP 99 dataset, NSL- KDD dataset and UNSW- NB15 dataset. Comparisons were performed with SAVAERDNN model. Oversampling technique is used for data balancing, and the stacking architecture handles the issue of overtraining introduced due to oversampling. Findings: The proposed Stacking and Feature engineeringbased Semi-supervised (SFS) model presents a combined architecture that integrates data balancing, feature engineering and a stacking-based prediction model that balances data to reduce imbalance, reduces the data size, and also provides highly effective predictions. Results: indicate 2% increase in accuracy levels on the UNSW-NB15 dataset and 10% increase in accuracy levels in the NSL-KDD dataset. Novelty: The architecture has been designed in a domainspecific manner. Multiple intrusion detection datasets, each with different levels of imbalance, have been used to depict the generic nature of the SFS model. Keywords: Intrusion Detection; Data Imbalance; Stacking; Feature Engineering; Oversampling; Semi Supervised Learning

Sampling-attention deep learning network with transfer learning for large-scale urban point cloud semantic segmentation

Targeting the development of smart cities to facilitate the significant analysis of large-scale urban for construction and update. This research develops a new method named transfer learning-based sampling-attention network (TSANet) to act on 3D urban point clouds for semantic segmentation. The main contributions of this research are a segmentation model and a transfer learning protocol, where the segmentation model adopts the point downsampling–upsampling structure for time efficiency, the embedding method and an attention mechanism for point cloud feature processing, and the transfer learning protocol is employed to reduce the data requirements and labeling efforts by using prior knowledge for practical application. In addition, a focal loss is designed to assist the model for feature extraction and learning with handling data imbalance. To demonstrate the efficiency and effectiveness of the developed method, a realistic point cloud dataset of Cambridge and Birmingham cities is utilized as a case study. The results demonstrate that (1) the developed method has promising performance with overall accuracy (OA) of 0.9133 and Mean Intersection over Union (MIoU) of 0.5588; (2) the proposed transfer learning protocol contributes to the core model performance by fully combining accuracy and time efficiency, offering a 74.91% improvement in time efficiency; (3) the developed TSANet outperforms other state-of-the-art models, such as PointNet++ and DGCNN, by comparing the accuracy and time efficiency. Overall, the method developed in this research is capable of great potential as a practical application tool by presenting accurate, effective, and efficient results for semantic segmentation of large-scale 3D urban point clouds.

Chi2-MI: A hybrid feature selection based machine learning approach in diagnosis of chronic kidney disease

• Development of an intelligent diagnosis system to detect chronic kidney disease. • A hybrid wrapper feature selection method (Chi 2 -MI) has been proposed and applied. • Data pre-processing methods are adopted to prepare the dataset for the model. • Most impactful features based on correlation scores are selected to predict the CKD. • Extra tress classifier can diagnose CKD with 98% accuracy among 14 learning models. Early detection and characterization are considered crucial in treating and controlling the chronic renal disease. Because of the rising number of patients, the high risk of progression to end-stage renal disease, and the poor prognosis of morbidity and mortality, chronic kidney disease (CKD) is a significant burden on the healthcare system. Detecting CKD in its early stages is critical for saving millions of lives. The uniqueness of this study lies in developing a diagnosis system to detect chronic kidney disease using different Machine Learning (ML) algorithms with the support of a hybrid feature selection approach. This study exploited the 400 clinical data of CKD patients based on the dataset supplied by the University of California Irvine (UCI) available at their Machine Learning repository. Different data preparation techniques like encoding categorical features, missing values imputation, removing outlier factors, handling data imbalance, scaling data at the same level, and selecting relevant features are adopted to prepare the dataset for the prediction model. A hybrid Chi-squared test (Chi2) and Mutual Information (MI) based feature selection approach is proposed to remove redundant features, and a Pearson correlation matrix is also computed to consider the top important features for the prediction. Lastly, the Extra tress classifier can diagnose CKD with 98% accuracy and a 2% true negative rate without data leakage out of 14 machine learning models. On the other hand, the Bagging classifier performed worst with only 60% accuracy.

Dealing with imbalanced regression problem for large dataset using scalable Artificial Neural Network

Although most of the machine learning and deep learning model expects the nature of data to have Gaussian representation for better predictive capability, in reality, this assumption does not hold true of late. In this paper, we explore the efficacy of an Artificial Neural Network (ANN) model using the K fold cross-validation strategy to deal with an imbalanced dataset. The objective of our work is to propose a robust scalable shallow ANN model which can handle data imbalance problems and train a large amount of data in less time. We demonstrate our model on the Redshift dataset from Astronomy where we consider the entire span of (0<z<7) which is highly skewed. The logarithmic transformation has been used here on redshift to tackle the issue of data imbalance. We have applied error metrics such as Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Normalized Median Absolute Deviation (NMAD) to measure the effectiveness of the ANN model and it shows the proposed model is feasible and reliable with a satisfactory RMSE (.89) and NMAD (.05) for a highly imbalanced dataset.

AD-CovNet: An exploratory analysis using a hybrid deep learning model to handle data imbalance, predict fatality, and risk factors in Alzheimer's patients with COVID-19

Alzheimer's disease (AD) is the leading cause of dementia globally, with a growing morbidity burden that may exceed diagnosis and management capabilities. The situation worsens when AD patient fatalities are exposed to COVID-19. Because of differences in clinical features and patient condition, a patient's recovery from COVID-19 with or without AD varies greatly. Thus, this situation stimulates a spectrum of imbalanced data. The inclusion of different features in the class imbalance offers substantial problems for developing of a classification framework. This study proposes a framework to handle class imbalance and select the most suitable and representative datasets for the hybrid model. Under this framework, various state-of-the-art resample techniques were utilized to balance the datasets, and three sets of data were finally selected. We developed a novel hybrid deep learning model AD-CovNet using Long Short-Term Memory (LSTM) and Multi-layer Perceptron (MLP) algorithms that delineate three unique datasets of COVID-19 and AD-COVID-19 patient fatality predictions. This proposed model achieved 97% accuracy, 97% precision, 97% recall, and 97% F1-score for Dataset I; 97% accuracy, 97% precision, 96% recall, and 96% F1-score for Dataset II; and 86% accuracy, 88% precision, 88% recall, and 86% F1-score for Dataset III. In addition, AdaBoost, XGBoost, and Random Forest models were utilized to evaluate the risk factors associated with AD-COVID-19 patients, and the outcome outperformed diagnostic performance. The risk factors predicted by the models showed significant clinical importance and relevance to mortality. Furthermore, the proposed hybrid model's performance was evaluated using a statistical significance test and compared to previously published works. Overall, the uniqueness of the large dataset, the effectiveness of the deep learning architecture, and the accuracy and performance of the hybrid model showcase the first cohesive work that can formulate better predictions and help in clinical decision-making.

Prototype-based classifier learning for long-tailed visual recognition

In this paper, we tackle the long-tailed visual recognition problem from the categorical prototype perspective by proposing a prototype-based classifier learning (PCL) method. Specifically, thanks to the generalization ability and robustness, categorical prototypes reveal their advantages of representing the category semantics. Coupled with their class-balance characteristic, categorical prototypes also show potential for handling data imbalance. In our PCL, we propose to generate the categorical classifiers based on the prototypes by performing a learnable mapping function. To further alleviate the impact of imbalance on classifier generation, two kinds of classifier calibration approaches are designed from both prototype-level and example-level aspects. Extensive experiments on five benchmark datasets, including the large-scale iNaturalist, Places-LT, and ImageNet-LT, justify that the proposed PCL can outperform state-of-the-arts. Furthermore, validation experiments can demonstrate the effectiveness of tailored designs in PCL for long-tailed problems.

A novel approach for fraudulent reviewer detection based on weighted topic modelling and nearest neighbors with asymmetric Kullback–Leibler divergence

The task of detecting fraudulent reviewers is of great importance to E-commerce platforms. Existing research has invested much effort into developing comprehensive features and advanced techniques to detect fraudulent reviewers. However, most of these studies have ignored the data imbalance problem inherent in fraudulent reviewer detection: non-fraudulent reviewers are the majority, while fraudulent reviewers are the minority in real practice. To fill this gap, we propose a novel approach called ImDetector to detect fraudulent reviewers while handling data imbalance based on weighted latent Dirichlet allocation (LDA) and Kullback–Leibler (KL) divergence. Specifically, we develop a weighted LDA model to extract the latent topics of reviewers distributed on the review features. Asymmetric KL divergence is adopted to make the similarity measure between reviewers biased toward the fraudulent minority when using the K-nearest-neighbor for classification. By mapping the reviewers to the latent topics of features derived from the weighted LDA model and measuring the similarities between reviewers using asymmetric KL divergence, the data imbalance problem in fraudulent reviewer detection is alleviated. Extensive experiments on the Yelp.com dataset demonstrate that the proposed ImDetector approach is superior to the state-of-the-art techniques used for fraudulent reviewer detection. We also explain the experimental results and present the managerial implications of this paper.

How to Handle Data Imbalance and Feature Selection Problems in CNN-Based Stock Price Forecasting

Stock market forecasting is a time series problem that aims to predict possible future prices or directions of an index/stock. The stock data contains high uncertainty and is influenced by too many factors; hence it isn’t easy to achieve the goal by traditional time series methods. In literature, the convolutional neural networks (CNN) models were used for stock market forecasting and gave successful results. But, data imbalance due to labeling and feature selection problems were seen when considering these models. Hence, this study proposed a new rule-based labeling algorithm and a new feature selection approach to solve the issues. In addition, a CNN-based model, which was presented to predict the next day’s trade action of stocks in the Dow30 index, was constructed to check the effectiveness of the data labeling and the feature selection approach. Different image-based input variable sets were created using technical indicators, gold, and oil price data to feed the CNN model. The prediction performance of CNN models was compared with other studies in the literature. The experimental results showed that the CNN prediction model, which uses the proposed feature selection and labeling approaches in this study, performs 3-22% higher accuracy than the CNN-based models taking part in other studies. Also, the labeling approach proposed is more successful than Chen and Huang’s data weighting approach to solve the stock data imbalance problem. This algorithm reduced the ratio between labeled data from 15 times to 1.8 times.

Balancing data for generalizable machine learning to predict glass-forming ability of ternary alloys

Machine Learning has thrived on the emergence of data-driven materials science. However, the materials datasets acquired at existing research efforts have significant imbalance issues. This paper investigated the data imbalance for the glass-forming ability of ternary alloy systems, which consists of abundant, low-fidelity high-throughput data, and sparse, high-fidelity traditional experimental data. We demonstrated a new method to handle the data imbalance and trained artificial neural network (ANN) models on the original vs. balanced datasets. The ANN model trained on the balanced dataset solved the overfitting issue suffered by the model trained on the original dataset. More importantly, the generalizability in predicting the new alloy system was improved in the data-balanced model, evidenced by the leave-one-alloy-system-out validation. Our work highlights the importance of handling data imbalance in material datasets to solve the overfitting issues of machine learning models and further enhance generalizability in predicting the characteristics of the new material systems.

FEATURE SUB-SPACING BASED STACKING FOR EFFECTIVE IMBALANCE HANDLING IN SENSITIVE DATA

Several real world classification applications suffer from an issue called data imbalance. Handling data imbalance is crucial in developing an effective classification system. This work presents an effective classifier ensemble model, Feature Sub-spacing Stacking Model (FSSM) that has been designed to operate on highly imbalanced, complex and sensitive data. The FSSM technique is based on creating subspace of features, to aid in the reduction of data complexity and also to handle data imbalance. First level trains models based on these features, which is followed by creating a stacking architecture. The second level stacking architecture trains on the predictions from the first level base models. This has enabled better and qualitative predictions. Experiments were conducted on bank data and also the NSL-KDD data. Results reveal highly effective performances compared to the existing models.

Addressing data imbalance problems in ligand-binding site prediction using a variational autoencoder and a convolutional neural network.

Since 2015, a fast growing number of deep learning-based methods have been proposed for protein-ligand binding site prediction and many have achieved promising performance. These methods, however, neglect the imbalanced nature of binding site prediction problems. Traditional data-based approaches for handling data imbalance employ linear interpolation of minority class samples. Such approaches may not be fully exploited by deep neural networks on downstream tasks. We present a novel technique for balancing input classes by developing a deep neural network-based variational autoencoder (VAE) that aims to learn important attributes of the minority classes concerning nonlinear combinations. After learning, the trained VAE was used to generate new minority class samples that were later added to the original data to create a balanced dataset. Finally, a convolutional neural network was used for classification, for which we assumed that the nonlinearity could be fully integrated. As a case study, we applied our method to the identification of FAD- and FMN-binding sites of electron transport proteins. Compared with the best classifiers that use traditional machine learning algorithms, our models obtained a great improvement on sensitivity while maintaining similar or higher levels of accuracy and specificity. We also demonstrate that our method is better than other data imbalance handling techniques, such as SMOTE, ADASYN, and class weight adjustment. Additionally, our models also outperform existing predictors in predicting the same binding types. Our method is general and can be applied to other data types for prediction problems with moderate-to-heavy data imbalances.

Handling imbalance in hierarchical classification problems using local classifiers approaches

The task of learning from imbalanced datasets has been widely investigated in the binary, multi-class and multi-label classification scenarios. Although this problem also affects hierarchical datasets, there are few work in the literature dealing with it. Meanwhile, the local classifier approaches are the most used techniques in the literature to deal with Hierarchical Classification problems. In this paper, we present new ways to handle data imbalance in hierarchical classification problems when using local classifiers approaches. We propose three different resampling schemas, according to the local classification approach: (1) Local Classifiers per Node; (2) Local Classifiers per Parent Node; and (3) Local Classifiers per Level. In order to define how imbalanced a certain hierarchical dataset is, we also propose three novel metrics to measure the imbalance in hierarchical datasets considering the different local classification approaches. The experimental evaluation in eight well-known datasets showed that the imbalance metrics can indeed measure the datasets imbalance and the proposed resampling schemas are able to improve the classification results when compared to baselines, state-of-the-art and related work approaches.

TGANs with Machine Learning Models in Automobile Insurance Fraud Detection and Comparative Study with Other Data Imbalance Techniques

A data-driven Fraud detection model for insurance business can be seen as a two-phase method. Phase I is data-preprocessing of a given dataset, in which, handling class imbalance is a major challenge. Phase II is that of classification using Machine Learning models. It is important to comprehend if there is any influence of the technique used in Phase I on the efficiency of the model used for Phase II. A natural query that intrigues one is whether there is a golden combination of a technique in Phase I and a specific model in Phase II for assured best performance of a Fraud Detection Model.In this work, we study a few techniques for handling data imbalance issue namely, SMOTE, MWMOTE, ADASYN and TGAN in combination with various classifier models like Random Forest (RF), Decision Trees (DT), Support Vector Machines (SVM), LightGBM, XGBoost and Gradient Boosting Machines (GBM). The study is conducted on a dataset for motor vehicle insurance fraud detection.We present a comparison of various combinations of data imbalance technique and classifier models. It is observed that the combination of TGAN in Phase I and GBM in Phase II gives the best performance. This combination performs best in terms of important metrics such as false positive rate, precision and specificity. We obtained the lowest false positive rate of 0.0011 and precision of 0.9988 which minimizes the most critical risk for the insurance company of falsely classifying a non-fraud claim as a fraud. Finally, the specificity of 0.9989 indicates that the model was also very good at predicting the non-fraudulent claim.

StackNet-DenVIS: a multi-layer perceptron stacked ensembling approach for COVID-19 detection using X-ray images.

The highly contagious nature of Coronavirus disease 2019 (Covid-19) resulted in a global pandemic. Due to the relatively slow and taxing nature of conventional testing for Covid-19, a faster method needs to be in place. The current researches have suggested that visible irregularities found in the chest X-ray of Covid-19 positive patients are indicative of the presence of the disease. Hence, Deep Learning and Image Classification techniques can be employed to learn from these irregularities, and classify accordingly with high accuracy. This research presents an approach to create a classifier model named StackNet-DenVIS which is designed to act as a screening process before conducting the existing swab tests. Using a novel approach, which incorporates Transfer Learning and Stacked Generalization, the model aims to lower the False Negative rate of classification compensating for the 30% False Negative rate of the swab tests. A dataset gathered from multiple reliable sources consisting of 9953 Chest X-rays (868 Covid and 9085 Non-Covid) was used. Also, this research demonstrates handling data imbalance using various techniques involving Generative Adversarial Networks and sampling techniques. The accuracy, sensitivity, and specificity obtained on our proposed model were 95.07%, 99.40% and 94.61% respectively. To the best of our knowledge, the combination of accuracy and false negative rate obtained by this paper outperforms the current implementations. We must also highlight that our proposed architecture also considers other types of viral pneumonia. Given the unprecedented sensitivity of our model we are optimistic it contributes to a better Covid-19 detection.

A proposed method for handling an imbalance data in classification of blood type based on Myers-Briggs type indicator

Blood type still leads to an assumption about its relation to some personality aspects. This study observes preprocessing methods for improving the classification accuracy of MBTI data to determine blood type. The training and testing data use 250 data from the MBTI questionnaire answers given by 250 respondents. The classification uses the k-Nearest Neighbor (k-NN) algorithm. Without preprocessing, k-NN results in about 32 % accuracy, so it needs some preprocessing to handle data imbalance before the classification. The proposed preprocessing consists of two-stage, the first stage is the unsupervised resample, and the second is the supervised resample. For the validation, it uses ten cross-validations. The result of k-Nearest Neighbor classification after using these proposed preprocessing stages has finally increased the accuracy, F-score, and recall significantly.

Applying machine learning in motor activity time series of depressed bipolar and unipolar patients compared to healthy controls.

Current practice of assessing mood episodes in affective disorders largely depends on subjective observations combined with semi-structured clinical rating scales. Motor activity is an objective observation of the inner physiological state expressed in behavior patterns. Alterations of motor activity are essential features of bipolar and unipolar depression. The aim was to investigate if objective measures of motor activity can aid existing diagnostic practice, by applying machine-learning techniques to analyze activity patterns in depressed patients and healthy controls. Random Forrest, Deep Neural Network and Convolutional Neural Network algorithms were used to analyze 14 days of actigraph recorded motor activity from 23 depressed patients and 32 healthy controls. Statistical features analyzed in the dataset were mean activity, standard deviation of mean activity and proportion of zero activity. Various techniques to handle data imbalance were applied, and to ensure generalizability and avoid overfitting a Leave-One-User-Out validation strategy was utilized. All outcomes reports as measures of accuracy for binary tests. A Deep Neural Network combined with SMOTE class balancing technique performed a cut above the rest with a true positive rate of 0.82 (sensitivity) and a true negative rate of 0.84 (specificity). Accuracy was 0.84 and the Matthews Correlation Coefficient 0.65. Misclassifications appear related to data overlapping among the classes, so an appropriate future approach will be to compare mood states intra-individualistically. In summary, machine-learning techniques present promising abilities in discriminating between depressed patients and healthy controls in motor activity time series.

Mining Massive E-Health Data Streams for IoMT Enabled Healthcare Systems.

With the increasing popularity of the Internet-of-Medical-Things (IoMT) and smart devices, huge volumes of data streams have been generated. This study aims to address the concept drift, which is a major challenge in the processing of voluminous data streams. Concept drift refers to overtime change in data distribution. It may occur in the medical domain, for example the medical sensors measuring for general healthcare or rehabilitation, which may switch their roles for ICU emergency operations when required. Detecting concept drifts becomes trickier when the class distributions in data are skewed, which is often true for medical sensors e-health data. Reactive Drift Detection Method (RDDM) is an efficient method for detecting long concepts. However, RDDM has a high error rate, and it does not handle class imbalance. We propose an Enhanced Reactive Drift Detection Method (ERDDM), which systematically generates strategies to handle concept drift with class imbalance in data streams. We conducted experiments to compare ERDDM with three contemporary techniques in terms of prediction error, drift detection delay, latency, and ability to handle data imbalance. The experimentation was done in Massive Online Analysis (MOA) on 48 synthetic datasets customized to possess the capabilities of data streams. ERDDM can handle abrupt and gradual drifts and performs better than all benchmarks in almost all experiments.

Novel Volumetric Sub-region Segmentation in Brain Tumors.

A novel deep learning based model called Multi-Planar Spatial Convolutional Neural Network (MPS-CNN) is proposed for effective, automated segmentation of different sub-regions viz. peritumoral edema (ED), necrotic core (NCR), enhancing and non-enhancing tumor core (ET/NET), from multi-modal MR images of the brain. An encoder-decoder type CNN model is designed for pixel-wise segmentation of the tumor along three anatomical planes (axial, sagittal, and coronal) at the slice level. These are then combined, by incorporating a consensus fusion strategy with a fully connected Conditional Random Field (CRF) based post-refinement, to produce the final volumetric segmentation of the tumor and its constituent sub-regions. Concepts, such as spatial-pooling and unpooling are used to preserve the spatial locations of the edge pixels, for reducing segmentation error around the boundaries. A new aggregated loss function is also developed for effectively handling data imbalance. The MPS-CNN is trained and validated on the recent Multimodal Brain Tumor Segmentation Challenge (BraTS) 2018 dataset. The Dice scores obtained for the validation set for whole tumor (WT :NCR/NE +ET +ED), tumor core (TC:NCR/NET +ET), and enhancing tumor (ET) are 0.90216, 0.87247, and 0.82445. The proposed MPS-CNN is found to perform the best (based on leaderboard scores) for ET and TC segmentation tasks, in terms of both the quantitative measures (viz. Dice and Hausdorff). In case of the WT segmentation it also achieved the second highest accuracy, with a score which was only 1% less than that of the best performing method.

Discriminative Fast Hierarchical Learning for Multiclass Image Classification.

In this article, a discriminative fast hierarchical learning algorithm is developed for supporting multiclass image classification, where a visual tree is seamlessly integrated with multitask learning to achieve fast training of the tree classifier hierarchically (i.e., a set of structural node classifiers over the visual tree). By partitioning a large number of categories hierarchically in a coarse-to-fine fashion, a visual tree is first constructed and further used to handle data imbalance and identify the interrelated learning tasks automatically (e.g., the tasks for learning the node classifiers for the sibling child nodes under the same parent node are strongly interrelated), and a multitask SVM classifier is trained for each nonleaf node to achieve more effective separation of its sibling child nodes at the next level of the visual tree. Both the internode visual similarities and the interlevel visual correlations are utilized to train more discriminative multitask SVM classifiers and control the interlevel error propagation effectively, and a stochastic gradient descent (SGD) algorithm is developed for learning such multitask SVM classifiers with higher efficiency. Our experimental results have demonstrated that our fast hierarchical learning algorithm can achieve very competitive results on both the classification accuracy rates and the computational efficiency.

Real‐time anomaly detection using parallelized intrusion detection architecture for streaming data

SummaryHigh usage levels of networking technologies has resulted in large amounts of data being generated. This in‐turn has lured several fraudsters, whose anomalous behaviors create undesired consequences to legitimate users. This paper proposes an Adaptive Parallelized Intrusion Detection (APID) architecture to handle the hugeness and data imbalance associated with streaming data. The architecture is composed of a feature selection strategy to reduce data size, an effective data segregation mechanism to handle data imbalance and a heterogeneous ensemble and a heuristic combiner mechanism to provide effective predictions. Adaptivity is incorporated by the reinforcement mechanism that retrains the model based on false predictions given by the model. The proposed APID architecture is generic; hence, it supports heterogeneous models and can also incorporate any number of machine learning models. Hence, it becomes flexible to adapt the model to data pertaining to any domain. Experiments were performed with KDD CUP 99, NSL‐KDD, and Koyoto 2006 datasets. Comparisons performed with recent works in literature indicates anomaly detection rates between 98% to 99% exhibiting the effectiveness of the proposed model.