Multi-class Data Research Articles

Survey on Highly Imbalanced Multi-class Data

Machine learning technology has a massive impact on society because it offers solutions to solve many complicated problems like classification, clustering analysis, and predictions, especially during the COVID-19 pandemic. Data distribution in machine learning has been an essential aspect in providing unbiased solutions. From the earliest literatures published on highly imbalanced data until recently, machine learning research has focused mostly on binary classification data problems. Research on highly imbalanced multi-class data is still greatly unexplored when the need for better analysis and predictions in handling Big Data is required. This study focuses on reviews related to the models or techniques in handling highly imbalanced multi-class data, along with their strengths and weaknesses and related domains. Furthermore, the paper uses the statistical method to explore a case study with a severely imbalanced dataset. This article aims to (1) understand the trend of highly imbalanced multi-class data through analysis of related literatures; (2) analyze the previous and current methods of handling highly imbalanced multi-class data; (3) construct a framework of highly imbalanced multi-class data. The chosen highly imbalanced multi-class dataset analysis will also be performed and adapted to the current methods or techniques in machine learning, followed by discussions on open challenges and the future direction of highly imbalanced multi-class data. Finally, for highly imbalanced multi-class data, this paper presents a novel framework. We hope this research can provide insights on the potential development of better methods or techniques to handle and manipulate highly imbalanced multi-class data.

Novel Oversampling Algorithm for Handling Imbalanced Data Classification

In the current age, the attention of researchers is immersed by numerous imbalanced data applications. These application areas are intrusion detection in security, fraud recognition in finance, medical applications dealing with disease diagnosis pilfering in electricity, and many more. Imbalanced data applications are categorized into two types: binary and multiclass data imbalance. Unequal data distribution among data diverts classification performance metrics towards the majority data instance class and ignores the minority data, instance class. Data imbalance leads to an increase in the classification error rate. Random Forest Classification (RFC) is best suitable technique to deal with imbalanced datasets. This paper proposes the novel oversampling rate calculation algorithm as Improvised Dynamic Binary-Multiclass Imbalanced Oversampling Rate (IDBMORate). Experimentation analysis of the proposed novel approach IDBMORate on Page-block (Binary) dataset shows that instances of positive class is increased from 559 to 1118 whereas negative instance class remains same as 4913. In case of referred multiclass dataset (Ecoli), IDBMORate produces the consistent result as minority classes (om, omL, imS, imL) instances are oversampled majority class instances remains unchanged. IDBMORate algorithm reduces the ignorance of minority class and oversamples its data without disturbing the size of the majority instance class. Thus, it reduces the overall computation cost and leads towards the improvisation of classification performance.

Multiclass data plane recovery using different recovery schemes in SDN: a simulation analysis

To provide dependable services SDN networks need to be resilient to link or switching node failures. This entails, when faults occur, ensuring differentiated types of recovery, according to carried traffic, to routing paths. However, the choice of the recovery scheme best suited to each traffic class is not direct, nor is obvious the impact of the combination of various recovery schemes, according to traffic classes. We explore the usage of different recovery schemes for traffic with distinct requirements Simulation analysis confirms that using different recovery schemes for distinct types of traffic does create differentiated effects in terms of traffic carried and bandwidth usage.

Water Wave Optimization with Deep Learning Driven Smart Grid Stability Prediction

Smart Grid (SG) technologies enable the acquisition of huge volumes of high dimension and multi-class data related to electric power grid operations through the integration of advanced metering infrastructures, control systems, and communication technologies. In SGs, user demand data is gathered and examined over the present supply criteria whereas the expenses are then informed to the clients so that they can decide about electricity consumption. Since the entire procedure is valued on the basis of time, it is essential to perform adaptive estimation of the SG’s stability. Recent advancements in Machine Learning (ML) and Deep Learning (DL) models enable the designing of effective stability prediction models in SGs. In this background, the current study introduces a novel Water Wave Optimization with Optimal Deep Learning Driven Smart Grid Stability Prediction (WWOODL-SGSP) model. The aim of the presented WWOODL-SGSP model is to predict the stability level of SGs in a proficient manner. To attain this, the proposed WWOODL-SGSP model initially applies normalization process to scale the data to a uniform level. Then, WWO algorithm is applied to choose an optimal subset of features from the pre-processed data. Next, Deep Belief Network (DBN) model is followed to predict the stability level of SGs. Finally, Slime Mold Algorithm (SMA) is exploited to fine tune the hyperparameters involved in DBN model. In order to validate the enhanced performance of the proposed WWOODL-SGSP model, a wide range of experimental analyses was performed. The simulation results confirm the enhanced predictive results of WWOODL-SGSP model over other recent approaches.

TimesVector-Web: A Web Service for Analysing Time Course Transcriptome Data with Multiple Conditions.

From time course gene expression data, we may identify genes that modulate in a certain pattern across time. Such patterns are advantageous to investigate the transcriptomic response to a certain condition. Especially, it is of interest to compare two or more conditions to detect gene expression patterns that significantly differ between them. Time course analysis can become difficult using traditional differentially expressed gene (DEG) analysis methods since they are based on pair-wise sample comparison instead of a series of time points. Most importantly, the related tools are mostly available as local Software, requiring technical expertise. Here, we present TimesVector-web, which is an easy to use web service for analysing time course gene expression data with multiple conditions. The web-service was developed to (1) alleviate the burden for analyzing multi-class time course data and (2) provide downstream analysis on the results for biological interpretation including TF, miRNA target, gene ontology and pathway analysis. TimesVector-web was validated using three case studies that use both microarray and RNA-seq time course data and showed that the results captured important biological findings from the original studies.

Medical QoS provisioning for multi-class data in coexisting wireless body area networks

Medical QoS provisioning for multi-class data in coexisting wireless body area networks

An oversampling method for multi-class imbalanced data based on composite weights.

To solve the oversampling problem of multi-class small samples and to improve their classification accuracy, we develop an oversampling method based on classification ranking and weight setting. The designed oversampling algorithm sorts the data within each class of dataset according to the distance from original data to the hyperplane. Furthermore, iterative sampling is performed within the class and inter-class sampling is adopted at the boundaries of adjacent classes according to the sampling weight composed of data density and data sorting. Finally, information assignment is performed on all newly generated sampling data. The training and testing experiments of the algorithm are conducted by using the UCI imbalanced datasets, and the established composite metrics are used to evaluate the performance of the proposed algorithm and other algorithms in comprehensive evaluation method. The results show that the proposed algorithm makes the multi-class imbalanced data balanced in terms of quantity, and the newly generated data maintain the distribution characteristics and information properties of the original samples. Moreover, compared with other algorithms such as SMOTE and SVMOM, the proposed algorithm has reached a higher classification accuracy of about 90%. It is concluded that this algorithm has high practicability and general characteristics for imbalanced multi-class samples.

Heart sound classification using signal processing and machine learning algorithms

According to global statistics and the world health organization (WHO), about 17.5 million people die each year from cardiovascular disease. In this paper, the heart sounds gathered by a stethoscope are analyzed to diagnose several diseases caused by heart failure. This research’s primary process is to identify and classify the data related to the heart sounds categorized in four general groups of S1 to S4. The sounds S1 and S2 are considered as the heart’s normal sounds, and the sounds S3 and S4 are the abnormal sounds of the heart (heart murmurs), each expressing a specific type of heart disease. In this regard, the desired features are first extracted after retrieving the data by signal processing algorithms. In the next step, feature selection algorithms are used to select the compelling features to reduce the problem’s dimensions and obtain the optimal answer faster. While the existing algorithms in the literature classify the sound into two groups of normal and abnormal, in the final section, some of the most popular classification algorithms are utilized to classify the type of sound into three classes of normal, S3 and S4 categories. The proposed methodology obtained an accuracy rate of 87.5% and 95% for multiclass data (3 classes) and 98% for binary classification (normal vs. abnormal) problems.

Multi-class Bayesian support vector data description with anomalies

Multi-class Bayesian support vector data description with anomalies

Classification of pulse flours using near-infrared hyperspectral imaging

With increasing consumer interest in healthy food formulations made from pulse flours, the research gap in characterizing them based on their pulse type and milling technique has come to limelight. To this end, the feasibility of using hyperspectral imaging in the visible near infrared (Vis-NIR) (400–1000 nm) and short wave infrared (SWIR) (1000–2500 nm) regions to classify pulse flours (viz. chickpea, yellow pea, navy bean and green lentil) based on the pulse type and milling methods was investigated. Unsupervised and supervised classification models were developed using unsupervised principal component analysis (PCA) and supervised partial least squares discriminant analysis (PLS-DA). Supervised classification for pulse flour type demonstrated 100% accuracy in the Vis-NIR wavelength range of 530–700 nm, which primarily exploited the color attributes of the flour samples. For milling method based classification, PLS-DA models developed using SWIR regions of 1370–1500 nm and 1700–2000 nm played a significant role in discriminating flour samples yielding 95% classification accuracy. These regions are associated with the O–H and N–H overtones of the proteins found in flour samples. Conclusively, hyperspectral imaging in the range of 400–2500 nm combined with multivariate data classification methods can reliably be used by the food industry for characterization of pulse flours.

Evolutionary Mahalanobis Distance-Based Oversampling for Multi-Class Imbalanced Data Classification.

The number of sensing data are often imbalanced across data classes, for which oversampling on the minority class is an effective remedy. In this paper, an effective oversampling method called evolutionary Mahalanobis distance oversampling (EMDO) is proposed for multi-class imbalanced data classification. EMDO utilizes a set of ellipsoids to approximate the decision regions of the minority class. Furthermore, multi-objective particle swarm optimization (MOPSO) is integrated with the Gustafson–Kessel algorithm in EMDO to learn the size, center, and orientation of every ellipsoid. Synthetic minority samples are generated based on Mahalanobis distance within every ellipsoid. The number of synthetic minority samples generated by EMDO in every ellipsoid is determined based on the density of minority samples in every ellipsoid. The results of computer simulations conducted herein indicate that EMDO outperforms most of the widely used oversampling schemes.

Multiclass Energy Packet Networks with finite capacity energy queues

Energy Packet Network (EPN) consists of a queueing network formed by N blocks, where each of them is formed by one data queue, that handles the workload, and one energy queue, that handles packets of energy.We study an EPN model where the energy packets start the transfer. In this model, energy packets are sent to the data queue of the same block. An energy packet routes one workload packet to the next block if the data queue is not empty, and it is lost otherwise.We assume that the energy queues have a finite buffer size and if an energy packet arrives to the system when the buffer is full, jump-over blocking (JOB) is performed, and therefore with some probability it is sent to the data queue and it is lost otherwise.We first provide a value of the jump-over blocking probability such that the steady-state probability distribution of packets in the queues admits a product form solution. The product form is established for multiserver and multiclass data packet queues under FCFS, preemptive LCFS and PS discipline. Moreover, in the case of a directed tree queueing network, we show that the number of data packets in each subtree decreases as the JOB probability increases for each block.

ATR-FTIR spectroscopy-based differentiation of hepatitis C and dengue infection in human freeze-dried sera

In the present study, we evaluated attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy for the differential diagnosis hepatitis C virus (HCV) and Dengue virus (DENV) infected freeze-dried sera along with healthy freeze-dried sera. Spectral signatures were recorded from 50 freeze-dried sera samples of each sample type. To achieve improved classification accurateness between the three sample types, multivariate data classification algorithms principal component analysis (PCA) and principal component regression (PCR) were used. Outcomes show 96 % inter-class variation as depicted by PCA and a classification accuracy of 99.2 % using PCR as shown by the value of R2; therefore, we propose that ATR-FTIR coupled with chemo metric data classification algorithms has considerable potential to distinguishHCV and DENV infected freeze-dried sera samples.

A Joint MLE Approach to Large-Scale Structured Latent Attribute Analysis

Structured latent attribute models (SLAMs) are a family of discrete latent variable models widely used in education, psychology, and epidemiology to model multivariate categorical data. A SLAM assumes that multiple discrete latent attributes explain the dependence of observed variables in a highly structured fashion. Usually, the maximum marginal likelihood estimation approach is adopted for SLAMs, treating the latent attributes as random effects. The increasing scope of modern assessment data involves large numbers of observed variables and high-dimensional latent attributes. This poses challenges to classical estimation methods and requires new methodology and understanding of latent variable modeling. Motivated by this, we consider the joint maximum likelihood estimation (MLE) approach to SLAMs, treating latent attributes as fixed unknown parameters. We investigate estimability, consistency, and computation in the regime where sample size, number of variables, and number of latent attributes all can diverge. We establish the statistical consistency of the joint MLE and propose efficient algorithms that scale well to large-scale data for several popular SLAMs. Simulation studies demonstrate the superior empirical performance of the proposed methods. An application to real data from an international educational assessment gives interpretable findings of cognitive diagnosis.

Multi-class feature selection by exploring reliable class correlation

High-dimensional multi-class data are ubiquitous in many domains and often have complicated but semantic relevance among classes. Preserving these semantic relationships in the process of feature selection is crucial to learn the nature of the objects under study. Many existing similarity preserving-based feature selection methods employ fixed similarities learned from the original feature space. However, the existence of redundant and irrelevant features in the original feature space may make the similarities unreliable to guide feature selection. In this paper, we propose to learn a dynamic similarity matrix P and a feature weighting matrix W with a two-stage method. P is calculated in the dimension reduction feature space according to W, then fixed and used to guide the next round of feature selection. Similarity matrix P and feature weighting matrix W will gradually become reliable after some repeat of the two stages. In this way, the proposed approach tries to capture reliable class correlation and select target features. Experiments on the data with noisy features and various types of public data show the attractiveness of the proposed method.

Towards a Hybrid Deep Learning Model for Anomalous Activities Detection in Internet of Things Networks

The tremendous number of Internet of Things (IoT) applications, with their ubiquity, has provided us with unprecedented productivity and simplified our daily life. At the same time, the insecurity of these technologies ensures that our daily lives are surrounded by vulnerable computers, allowing for the launch of multiple attacks via large-scale botnets through the IoT. These attacks have been successful in achieving their heinous objectives. A strong identification strategy is essential to keep devices secured. This paper proposes and implements a model for anomaly-based intrusion detection in IoT networks that uses a convolutional neural network (CNN) and gated recurrent unit (GRU) to detect and classify binary and multiclass IoT network data. The proposed model is validated using the BoT-IoT, IoT Network Intrusion, MQTT-IoT-IDS2020, and IoT-23 intrusion detection datasets. Our proposed binary and multiclass classification model achieved an exceptionally high level of accuracy, precision, recall, and F1 score.

A graphical approach for multiclass classification and for correcting the labeling errors in mislabeled training data

Multiclass data classification, where the goal is to segment data into classes, is an important task in machine learning. However, the task is challenging due to reasons including the scarcity of labeled training data; in fact, most machine learning algorithms require a large amount of labeled examples to perform well. Moreover, the accuracy of a classifier can be dependent on the accuracy of the training labels which can be corrupted. In this paper, we present an efficient and unconditionally stable semi-supervised graph-based method for multiclass data classification which requires considerably less labeled training data to accurately classify a data set compared to current techniques, due to properties such as the embedding of data into a similarity graph. In particular, it performs very well and more accurately than current approaches in the common scenario of few labeled training elements. Morever, we show that the algorithm performs with good accuracy even with a large number of mislabeled examples and is also able to incorporate class size information. The proposed method uses a modified auction dynamics technique. Extensive experiments on benchmark datasets are performed and the results are compared to other methods.

Implementation of Multivariate Exponential Power Distribution in Discrimination and Classification of Psychological Data and Other applications

Recent advances have shown that some multivariate psychological data are deviating from usual normal assumption either in the tails or kurtosis. Thereby, allowing the call for modelling of such data using more robust elliptically contoured density which includes the normal distribution as a special case. This allowed more flexibility at the kurtosis and tail regions, which is better in handling non-normality in data analysis and also lower the cost of misclassification. The present study employed a robust model for such cases in the context of discrimination and classification of multivariate psychological disorder data using multivariate exponential distribution as an underlining model. Parameters were estimated using the method of maximum likelihood estimation and the discrimination and classification were based on the log likelihood ratio approach. The resulting models relied solidly on the shape parameter, which regulate the tails and the kurtosis, thereby allowed flexibility. This method enable us to lower the cost of misclassification. Some other areas of applications were also considered in the paper.

Multi Class Data Classification to Improve Accuracy in Sentiment Analysis using Machine Learning

Sentiment analysis means classifying a text into different emotional classes. These days most of the sentiment analysis techniques divide the text into either binary or ternary classification in this paper we are classifying the movie reviews into 5 classes. Multi class sentiment analysis is a technique which can be used to know the exact sentiment of a review not just polarity of a given textual statement from positive to negative. So that one can know the precise sentiment of a review . Multi class sentiment analysis has always been a challenging task as natural languages are difficult to represent mathematically. The number of features are also generally large which requires huge computational power so to reduce the number of features we will use parts-of-speech tagging using textblob to extract the important features. Sentiment analysis is done using machine learning, where it requires training data and testing data to train a model. Various kinds of models are trained and tested at last one model is selected based on its accuracy and confusion matrix. It is important to analyze the reviews in textual form because large amount of reviews is present all over the web. Analyzing textual reviews can help the firms that are trying to find out the response of their products in the market. In this paper sentiment analysis is demonstrated by analyzing the movie reviews, reviews are taken from IMDB website.

Equivalence class selection of categorical graphical models

Learning the structure of dependence relations between variables is a pervasive issue in the statistical literature. A directed acyclic graph (DAG) can represent a set of conditional independencies, but different DAGs may encode the same set of relations and are indistinguishable using observational data. Equivalent DAGs can be collected into classes, each represented by a partially directed graph known as essential graph (EG). Structure learning directly conducted on the EG space, rather than on the allied space of DAGs, leads to theoretical and computational benefits. Still, the majority of efforts has been dedicated to Gaussian data, with less attention to methods designed for multivariate categorical data. A Bayesian methodology for structure learning of categorical EGs is then proposed. Combining a constructive parameter prior elicitation with a graph-driven likelihood decomposition, a closed-form expression for the marginal likelihood of a categorical EG model is derived. Asymptotic properties are studied, and an MCMC sampler scheme developed for approximate posterior inference. The methodology is evaluated on both simulated scenarios and real data, with appreciable performance in comparison with state-of-the-art methods.