Automatic Classification Techniques Research Articles

Multi-Class Heart Abnormalities Detection Based on ECG Graph Using Transfer Learning Method

The heart is one of the vital organs in the circulatory system. Regular checkups are very important to prevent heart disease. The most basic examination is blood pressure then further examination is related to the evaluation of the electrical activity of the heart using an electrocardiogram (ECG). The ECG carries important information regarding various abnormalities of heart function. Several automated classification techniques have been proposed to facilitate diagnosis. However, not all digital ECG devices provide raw data for analysis. ECG classification method based on images can be an alternative in classification. Therefore, in this study, it is proposed to classify ECG based on signal images. The proposed classification method uses transfer learning with VGG, AlexNet, and DenseNet architectures. The method used for the classification of multi-class ECG consists of normal, PVC, Atrial Fibrilation, AFL, Bigeminy, LBBB, and APB. The simulation results generate the best accuracy of 92% and F1-score of 92%. Best performance is achieved using DenseNet architecture at 60 epochs. This study is expected to be a new reference technique in the classification of ECG signals.

Vision-based techniques for automatic marine plankton classification

Plankton are an important component of life on Earth. Since the 19th century, scientists have attempted to quantify species distributions using many techniques, such as direct counting, sizing, and classification with microscopes. Since then, extraordinary work has been performed regarding the development of plankton imaging systems, producing a massive backlog of images that await classification. Automatic image processing and classification approaches are opening new avenues for avoiding time-consuming manual procedures. While some algorithms have been adapted from many other applications for use with plankton, other exciting techniques have been developed exclusively for this issue. Achieving higher accuracy than that of human taxonomists is not yet possible, but an expeditious analysis is essential for discovering the world beyond plankton. Recent studies have shown the imminent development of real-time, in situ plankton image classification systems, which have only been slowed down by the complex implementations of algorithms on low-power processing hardware. This article compiles the techniques that have been proposed for classifying marine plankton, focusing on automatic methods that utilize image processing, from the beginnings of this field to the present day.

Automated classification ofChandraX-ray point sources using machine learning methods

ABSTRACTA large number of unidentified sources found by astronomical surveys and other observations necessitate the use of an automated classification technique based on machine learning (ML) methods. The aim of this paper is to find a suitable automated classifier to identify the point X-ray sources in the Chandra Source Catalogue (CSC) 2.0 in the categories of active galactic nuclei (AGN), X-ray emitting stars, young stellar objects (YSOs), high-mass X-ray binaries (HMXBs), low-mass X-ray binaries (LMXBs), ultra luminous X-ray sources (ULXs), cataclysmic variables (CVs), and pulsars. The catalogue consists of ≈317 000 sources, out of which we select 277 069 point sources based on the quality flags available in CSC 2.0. In order to identify unknown sources of CSC 2.0, we use MW features, such as magnitudes in optical/ultraviolet bands from Gaia-EDR3, Sloan Digital Sky Survey, and GALEX, and magnitudes in infrared bands from 2MASS, WISE, and MIPS-Spitzer, in addition to X-ray features (flux and variability) from CSC 2.0. We find the Light Gradient Boosted Machine, an advanced decision tree-based ML classification algorithm, suitable for our purpose and achieve 93 per cent precision, 93 per cent recall score, and 0.91 Mathew’s Correlation coefficient score. With the trained classifier, we identified 54 770 (14 066) sources with more than 3σ (4σ) confidence, out of which there are 32 600 (8574) AGNs, 16,148 (5,166) stars, 5,184 (208) YSOs, 439 (46) HMXBs, 197 (71) LMXBs, 50 (0) ULXs, 89 (1) CVs, and 63 (0) pulsars. This method can also be useful for identifying sources of other catalogues reliably.

CLASSIFICATION OF RICE USING CONVOLUTIONAL NEURAL NETWORK (CNN)

This paper describes the technique for automatic recognition and classification of different rice grain samples using neural network classifier. The Red Green Blue (RGB), Hue Saturation Intensity (HSI) and Hue Saturation Value (HSV) color models of the image were considered for extracting 18 color features. The classification was carried out using color and texture features separately. The color image was converted to Gray scale image and the Gray Level Co-occurrence Matrixes (GLCM) for four different directions was calculated. A total of eight texture features were calculated from the Co-occurrence matrices. Convolutional Neural Network (CNN) is used for the classification process. The classification accuracy with color features and texture features were compared. Result shows that the classification base on texture features outperform the color feature-based classification even with lesser number of features. It is found that Convolutional Neural network was able to classify two varieties of rice with 100% accuracy using texture features and the edge detection with Sobel and Canny edge detection of the fiber features in the food grain.

An Automated Classification Technique for COVID-19 Using Optimized Deep Learning Features

In 2020, COVID-19 started spreading throughout the world. This deadly infection was identified as a virus that may affect the lungs and, in severe cases, could be the cause of death. The polymerase chain reaction (PCR) test is commonly used to detect this virus through the nasal passage or throat. However, the PCR test exposes health workers to this deadly virus. To limit human exposure while detecting COVID-19, image processing techniques using deep learning have been successfully applied. In this paper, a strategy based on deep learning is employed to classify the COVID-19 virus. To extract features, two deep learning models have been used, the DenseNet201 and the SqueezeNet. Transfer learning is used in feature extraction, and models are fine-tuned. A publicly available computerized tomography (CT) scan dataset has been used in this study. The extracted features from the deep learning models are optimized using the Ant Colony Optimization algorithm. The proposed technique is validated through multiple evaluation parameters. Several classifiers have been employed to classify the optimized features. The cubic support vector machine (Cubic SVM) classifier shows superiority over other commonly used classifiers and attained an accuracy of 98.72%. The proposed technique achieves state-of-the-art accuracy, a sensitivity of 98.80%, and a specificity of 96.64%.

Qualitative Abnormalities of Peripheral Blood Smear Images Using Deep Learning Techniques

In recent years, Peripheral blood smear is a generic analysis to assess the person’s health status. Manual testing of Peripheral blood smear images are difficult, time-consuming and is subject to human intervention and visual error. This method encouraged for researchers to present algorithms and techniques to perform the peripheral blood smear analysis with the help of computer-assisted and decision-making techniques. Existing CAD based methods are lacks in attaining the accurate detection of abnormalities present in the images. In order to mitigate this issue Deep Convolution Neural Network (DCNN) based automatic classification technique is introduced with the classification of eight groups of peripheral blood cells such as basophil, eosinophil, lymphocyte, monocyte, neutrophil, erythroblast, platelet, myocyte, promyocyte and metamyocyte. The proposed DCNN model employs transfer learning approach and additionally it carries three stages such as pre-processing, feature extraction and classification. Initially the pre-processing steps are incorporated to eliminate noisy contents present in the image by using Histogram Equalization (HE). It is enclosed to improve an image contrast. In order to distinguish the dissimilar class and segmentation approach is carried out with the help of Fuzzy C-Means (FCM) model whereas its centroid point optimality method with Slap Swarm based optimization strategy. Moreover some specific set of Gray Level Co-occurrence Matrix (GLCM) features of the segmented images are extracted to augment the performance of proposed detection algorithm. Finally the extracted features are recorded by DCNN and the proposed classifier has the capability to extract their own features. Based on this the diverse set of classes are classified and distinguished from qualitative abnormalities found in the image.

Improved Accuracy of Brain Tumor Detection Using VGG16

Abstract: Brain tumor is a very serious problem of human life. In recent years, brain tumors have become one of the leading causes of death among people. It is difficult to identify the tumor itself. Direct detection and classification of brain tumors has the potential to achieve high efficiency and high levels of prognosis. However, it is well known that the accuracy of automatic identification and classification techniques varies from one technique to another and depends on the nature of the image. For the diagnosis and classification of brain tumors, MRI images have been very useful in recent years. MRI images allow us to diagnose brain tumors. This paper highlights the techniques of CNN and has worked upon VGG16 model.

Brain Tumor Detection using Convolution Neural Network

Brain tumors are a deadly disease with a life expectancy of only a few months in the most advanced stages. As a result, therapy planning is an important step in improving patients' quality of life. Various image techniques, such as Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and Ultrasound images, are commonly used to assess tumors in the brain, lung, liver, breast, prostate, and other organs. MRI images are used to diagnose brain tumors in particular in this work .However the massive amount of data generated by MRI scan thwarts manual classification of tumor vs non-tumor in a particular time. As a result, a reliable and automatic classification technique is required to reduce the human fatality rate. Deep Learning has sparked a lot of interest in recent years. It has been widely used in a variety of applications and has proven to be an effective machine learning technique for a variety of complicated issues. The use of Convolution Neural Networks (CNN) classification for automatic brain tumor detection is proposed in this paper. Small kernels are used to create the deeper architecture. When compared to all other state-of-the-art methodologies, experimental results demonstrate that CNN archives have a rate of 97.5 percent accuracy with little complexity.

A survey of software architectural change detection and categorization techniques

Software architecture is defined as the structural construction, design decisions implementation, evolution and knowledge sharing mechanisms of a system. Software architecture documentation help architects with decision making, guide developers during implementation, and preserve architectural decisions so that future caretakers are able to better understand an architect’s solution. Many modern-day software development teams are focusing more on architectural consistency of software design to better cope with the cost-time-efforts, continuous integration, software glitches, security backdoors, regulatory inspections, human values, and so on. Therefore, in order to better reflect the software design challenges, the development teams review the architectural design either on a regular basis or after completing certain milestones or releases. However, many studies have focused on architectural change detection and classification as the essential steps for reviewing design, discovering architectural tactics and knowledge, analyzing software stability, tracing and auditing software development history, recovering design decisions, generating design summary, and so on.In this paper, we survey state-of-the-art architectural change detection and categorization techniques and identify future research directions. To the best of our knowledge, our survey is the first comprehensive report on this area. However, in this survey, we compare available techniques using various quality attributes relevant to software architecture for different implementation levels and types. Moreover, our analysis shows that there is a lack of lightweight techniques (in terms of human intervention, algorithmic complexity, and frequency of usage) feasible to process hundreds and thousands of change revisions of a project. We also realize that rigorous focuses are required for capturing the design decision associativity of the architectural change detection techniques for practical use in the design review process. However, our survey on architectural change classification shows that existing automatic change classification techniques are not promising enough to use for real-world scenarios and reliable post analysis of causes of architectural change is not possible without manual intervention. There is also a lack of empirical data to construct an architectural change taxonomy, and further exploration in this direction would add much value to architectural change management.

Use of Laughter for the Detection of Parkinson's Disease: Feasibility Study for Clinical Decision Support Systems, Based on Speech Recognition and Automatic Classification Techniques.

Parkinson’s disease (PD) is an incurable neurodegenerative disorder which affects over 10 million people worldwide. Early detection and correct evaluation of the disease is critical for appropriate medication and to slow the advance of the symptoms. In this scenario, it is critical to develop clinical decision support systems contributing to an early, efficient, and reliable diagnosis of this illness. In this paper we present a feasibility study for a clinical decision support system for the diagnosis of PD based on the acoustic characteristics of laughter. Our decision support system is based on laugh analysis with speech recognition methods and automatic classification techniques. We evaluated different cepstral coefficients to identify laugh characteristics of healthy and ill subjects combined with machine learning classification models. The decision support system reached 83% accuracy rate with an AUC value of 0.86 for PD–healthy laughs classification in a database of 20,000 samples randomly generated from a pool of 120 laughs from healthy and PD subjects. Laughter could be employed for the efficient and reliable detection of PD; such a detection system can be achieved using speech recognition and automatic classification techniques; a clinical decision support system can be built using the above techniques. Significance: PD clinical decision support systems for the early detection of the disease will help to improve the efficiency of available and upcoming therapeutic treatments which, in turn, would improve life conditions of the affected people and would decrease costs and efforts in public and private healthcare systems.

An automated liver tumour segmentation and classification model by deep learning based approaches

ABSTRACT Liver cancer is regarded as one of the most common and leading causes of cancer death around the world. Automatic liver tumour segmentation and classification techniques are essential for assisting doctors in the tumour diagnosis process. As artificial intelligence progresses, powerful classification algorithms that can modify a variety of real-world applications are becoming accessible. Due to noise, non-homogeneity and the considerable appearance variability seen in tumour tissue, classifying liver tumours is a difficult undertaking. We have offered a new automatic liver tumour segmentation and classification methodology in this paper. The procedure for detecting liver tumours consists of two steps: first, mask-RCNN (Regions with Convolutional Neural Networks) segmentation of the liver part and then MSER (Maximally Stable Extremal Regions) tumour identification. We have used a hybrid Convolution Neural Network (CNN) model based on deep learning to perform the classification. The segmentation framework attempts to distinguish between normal and malignant tissue in the liver, while the classification method computes multi-class classification of the tumours found. The goal of this research is to come up with an unbiased forecast that is independent of human error. Our proposed method, on the other hand, nearly equals the top segmentation and classifying performance and provides the highest precision for lesion identification while keeping a high recall value. Our proposed approach correctly classifies identified liver tumours into three categories: hepatocellular carcinomas (HCC), malignant (other than HCCs) and benign or cyst with an average accuracy of 87.8%. The novelty of this paper lies in designing a mask-RCNN-based method for segmentation of the liver portion, implementing MSER to segment tumour lesions and using a hybrid CNN-based approach to categorise liver masses.

Effective Data Augmentation, Filters, and Automation Techniques for Automatic 12-Lead ECG Classification Using Deep Residual Neural Networks.

Automatic electrocardiogram (ECG) analysis plays a critical role in early detection and diagnosis of cardiac abnormalities and diseases. Data augmentation and automation strategies have been proposed to enhance the robustness of the machine and deep learning model for the classification of cardiac abnormalities. Here we propose 15 data augmentation and 6 filters, and an automation method using an end-to-end deep residual neural network (ResNet) model for automatic cardiac abnormalities detection from 12-lead ECG recordings. We evaluate the effectiveness of data augmentation/filtering and automation techniques using the proposed ResNet-based model on the China Physiological Signal Challenge (CPSC) dataset consisting of 9 diagnostic classes. The average F1 scores across 9 classes on the CPSC dataset trained with three data augmentation (baseline wander addition, dropout, and scaling) and a filter (sigmoid compression) were significantly higher than that without using augmentation/filters (baseline). The highest average F1 score with sigmoid compression method was significantly higher (relative improvement of 2.04 %) than the baseline while horizontal and vertical flipping augmentations were detrimental to the classification performance. Additionally, the results show that the random combination of four selected data augmentation and filter using the modified RandAugment technique provided a significantly higher average F1 score (relative improvement of 2.54 %) compared to the baseline. The proposed data augmentation, filters, and automation techniques provide an effective solution to improve the classification performance of the end-to-end deep learning model from ECG recordings without changing the model hyperparameters and structure.

A comparison of automated classification techniques for image processing in video internet of things

The counts of various types of white blood cells give vital information for identifying a variety of ailments utilizing the video internet of things (VIoT). This technique needs to be automated to save time and minimize counting errors. This paper aims to apply recent image processing techniques to identify and categorize white blood cells (WBCs) found in peripheral blood and to develop a system that would automatically detect and evaluate WBCs. The proposed method consists of steps such as segmentation scanning, feature extraction, and blood cell categorization. First, we segment the cell pictures, which entails grouping white blood cells into clusters. The second stage consists of scanning each input image and preparing the dataset. The third stage involves determining the texture and contour of a scanned picture. Finally, different machine learning algorithms are used to classify the results based on these criteria.

An automatic classification and early disease detection technique for herbs plant

Agriculture sector plays a major role for the growth of global economy. This is one of the reasons why disease detection in herbs plants is vital in the agricultural area, as disease in them are naturally occurring phenomenon. If sufficient precautions are not taken in this sector, it can have major consequences for plants, affecting quality standards, availability, and production. Hence the identification of non-pathogenic and non-pathogenic characteristics among plants will enrich the global plant production industry and the pharmaceutical industry. However, the current detection methodologies identify the type of herbs but which not suitable for real-time scenario because subject to prediction errors. This study proposes an automatic classification and early disease detection technique for herbs plant using hybrid soft computing techniques (CDD-H-HSC). First, a multi-swarm coyote optimization (MSCO) algorithm is used for disease segmentation which separate diseases area from input plant leaf image. Second, improved Chan-Vese snake optimization (ICVSO) algorithm is applied to optimize the features to select best optimal among multiple features which reduces the dimensionality problem. Then, a hybrid soft computing technique is developed i.e. fitness-distance balance deep neural network (FDB-DNN) classifier classifies the leaf and detect the disease. Finally, to evaluate the performance of proposed CDD-H-HSC technique with standard benchmark datasets. Then, the performance of proposed FDB-DNN classifier is compared with the existing state-of-art classifiers in terms of accuracy, precision, recall and F-measure.

BIR-CAT Optimization Technique for Automatic Segmentation and Classification of Brain Tumours on Pre- and Post-Operative MRI

BIR-CAT Optimization Technique for Automatic Segmentation and Classification of Brain Tumours on Pre- and Post-Operative MRI

Classification of Bacterial Images using Transfer Learning, Optimized Training and Resnet-50

Bacterial image analysis using traditional laboratory methods encounters bacterial recognition errors and requires extra experience and long processing time. Therefore, the automated classification technique of bacterial images is more useful than traditional visual observations for biologists because of their accurate classification, low cost, and fast diagnosis. In this study, a method to classify bacteria images by implementing the CNN deep learning method using Transfer Learning is proposed. This trained ResNet-50 is implemented as the CNN architecture. In the training of the classification layer, SGDM optimizer is used. The classification performance for is evaluated in using confusion matrix and four performance metrics: Accuracy, Precision, Recall and Fmeasure. The Confusion Matrix and all the performance metrics show is successful in classifying bacterial images.

Classification of Bacterial Images using Transfer Learning, Optimized Training and Resnet-50

Bacterial image analysis using traditional laboratory methods encounters bacterial recognition errors and requires extra experience and long processing time. Therefore, the automated classification technique of bacterial images is more useful than traditional visual observations for biologists because of their accurate classification, low cost, and fast diagnosis. In this study, a method to classify bacteria images by implementing the CNN deep learning method using Transfer Learning is proposed. This trained ResNet-50 is implemented as the CNN architecture. In the training of the classification layer, SGDM optimizer is used. The classification performance for is evaluated in using confusion matrix and four performance metrics: Accuracy, Precision, Recall and Fmeasure. The Confusion Matrix and all the performance metrics show is successful in classifying bacterial images.

Detection of extragalactic Ultra-compact dwarfs and Globular Clusters using Explainable AI techniques

Compact stellar systems such as Ultra-compact dwarfs (UCDs) and Globular Clusters (GCs) around galaxies are known to be the tracers of the merger events that have been forming these galaxies. Therefore, identifying such systems allows to study galaxies mass assembly, formation and evolution. However, in the lack of spectroscopic information detecting UCDs/GCs using imaging data is very uncertain. Here, we aim to train a machine learning model to separate these objects from the foreground stars and background galaxies using the multi-wavelength imaging data of the Fornax galaxy cluster in 6 filters, namely u, g, r, i, J and Ks. The classes of objects are highly imbalanced which is problematic for many automatic classification techniques. Hence, we employ Synthetic Minority Over-sampling to handle the imbalance of the training data. Then, we compare two classifiers, namely Localized Generalized Matrix Learning Vector Quantization (LGMLVQ) and Random Forest (RF). Both methods are able to identify UCDs/GCs with a precision and a recall of >93% and provide relevances that reflect the importance of each feature dimension for the classification. Both methods detect angular sizes as important markers for this classification problem. While it is astronomical expectation that color indices of u−i and i−Ks are the most important colors, our analysis shows that colors such as g−r are more informative, potentially because of higher signal-to-noise ratio. Besides the excellent performance the LGMLVQ method allows further interpretability by providing the feature importance for each individual class, class-wise representative samples and the possibility for non-linear visualization of the data as demonstrated in this contribution. We conclude that employing machine learning techniques to identify UCDs/GCs can lead to promising results. Especially transparent methods allow further investigation and analysis of importance of the measurements for the detection problem and provide tools for non-linear visualization of the data.

SABDM: A self‐attention based bidirectional‐RNN deep model for requirements classification

AbstractThe success of software depends upon functional and non‐functional requirements as both requirements are equally important in software development. However, the requirements engineering community still lacks in comprehensive understanding of functional and non‐functional requirements. In addition, the requirements in software documents are expressed in natural language and also intertwined with each other. Requirements classification is a crucial task that correctly extracts functional and non‐functional requirements and organizes them in specified categories. Automated classification of software requirements leads to reduced ambiguity, misunderstanding, and development cost. Most of the recent studies have used machine learning and deep learning techniques for automatic classification of requirements. However, there is one drawback of such techniques, that is, poor generalization. To address these problems, this research work proposes self‐attention based bidirectional LSTM deep model. This automated approach has used recurrent neural network, which handle long sequential natural language requirements statements and classify them into five classes such as capability, maintainability, performance, security, and usability. The proposed approach train and evaluate over pre‐labeled dataset comprised of 34 industrial requirements specifications and PROMISE dataset. Over this dataset, the proposed approach yields 95% of precision, 96% of recall, 96% of F‐measure, and 96% of accuracy. The proposed approach can be applied to wide variety of datasets with different domain. Furthermore, this paper applies pre‐processing techniques to improve the performance of the requirements classification model. The results of the proposed model are compared with existing baseline state‐of‐art techniques, and it is shown that the proposed model outperforms the baseline models in requirements classification.

Efficient Deep CNN Model for COVID-19 Classification

Coronavirus (COVID-19) infection was initially acknowledged as a global pandemic in Wuhan in China. World Health Organization (WHO) stated that the COVID-19 is an epidemic that causes a 3.4% death rate. Chest X-Ray (CXR) and Computerized Tomography (CT) screening of infected persons are essential in diagnosis applications. There are numerous ways to identify positive COVID-19 cases. One of the fundamental ways is radiology imaging through CXR, or CT images. The comparison of CT and CXR scans revealed that CT scans are more effective in the diagnosis process due to their high quality. Hence, automated classification techniques are required to facilitate the diagnosis process. Deep Learning (DL) is an effective tool that can be utilized for detection and classification this type of medical images. The deep Convolutional Neural Networks (CNNs) can learn and extract essential features from different medical image datasets. In this paper, a CNN architecture for automated COVID-19 detection from CXR and CT images is offered. Three activation functions as well as three optimizers are tested and compared for this task. The proposed architecture is built from scratch and the COVID-19 image datasets are directly fed to train it. The performance is tested and investigated on the CT and CXR datasets. Three activation functions: Tanh, Sigmoid, and ReLU are compared using a constant learning rate and different batch sizes. Different optimizers are studied with different batch sizes and a constant learning rate. Finally, a comparison between different combinations of activation functions and optimizers is presented, and the optimal configuration is determined. Hence, the main objective is to improve the detection accuracy of COVID-19 from CXR and CT images using DL by employing CNNs to classify medical COVID-19 images in an early stage. The proposed model achieves a classification accuracy of 91.67% on CXR image dataset, and a classification accuracy of 100% on CT dataset with training times of 58 min and 46 min on CXR and CT datasets, respectively. The best results are obtained using the ReLU activation function combined with the SGDM optimizer at a learning rate of 10−5 and a minibatch size of 16.