High Classification Rate Research Articles

Using machine learning to study the population life quality: methodological aspects

Assessment of the population life quality is an important and relevant sociological task. Machine learning as a classification tool of social network users’ digital traces makes it possible to create a base to calculate subjective life quality index. The article consistently reviews all stages of the machine learning algorithms application to assess the life quality of the population of the regions of the Russian Federation and the issues of improving neural network accuracy. To train the neural network the authors formed a set of marked-up data extracted from regional communities of the social network “VKontakte”. Various approaches to text vectorisation, publicly available neural network models pre-trained on large Russian-language text corpora, as well as metrics for evaluating the algorithms results were analysed. Computational experiments with different algorithms were carried out, according to the results of which the Rubert-tiny algorithm was selected due to its high learning and classification rate. During the model parameters adjustment, the accuracy of f1-macro 0.545 was achieved. Computational experiments were carried out using Python scripts.Typical errors that a neural network makes in the process of automatic content classification were considered. The results of the study can be used to calculate the online activity index in the VKontakte social network of users from various Russian regions, on the basis of which the subjective life quality index will be calculated in the future. Improving the neural network accuracy will make it possible to obtain more reliable data for assessing the life quality in Russian regions based on users’ digital traces.

Preliminary Classification of Selected Farmland Habitats in Ireland Using Deep Neural Networks.

Ireland has a wide variety of farmlands that includes arable fields, grassland, hedgerows, streams, lakes, rivers, and native woodlands. Traditional methods of habitat identification rely on field surveys, which are resource intensive, therefore there is a strong need for digital methods to improve the speed and efficiency of identification and differentiation of farmland habitats. This is challenging because of the large number of subcategories having nearly indistinguishable features within the habitat classes. Heterogeneity among sites within the same habitat class is another problem. Therefore, this research work presents a preliminary technique for accurate farmland classification using stacked ensemble deep convolutional neural networks (DNNs). The proposed approach has been validated on a high-resolution dataset collected using drones. The image samples were manually labelled by the experts in the area before providing them to the DNNs for training purposes. Three pre-trained DNNs customized using the transfer learning approach are used as the base learners. The predicted features derived from the base learners were then used to train a DNN based meta-learner to achieve high classification rates. We analyse the obtained results in terms of convergence rate, confusion matrices, and ROC curves. This is a preliminary work and further research is needed to establish a standard technique.

Fault Location Identifications in HV Transmission Networks and Different MV Wind Farms Using Nonintrusive Monitoring Techniques

The transmission-line fault localization of high-voltage (HV) transmission networks and different medium-voltage wind farms is implemented using a global time–frequency analysis algorithm, energy-spectrum-based hyperbolic S-transform (HS), to extract the potential power signatures from monitoring nonlinear and nonstationary fault signals on HV power utility. An energy concentration algorithm transforms each HS coefficient into useful features to quantify various faulty events and reduce recognition algorithms’ inputs. Furthermore, the multiclass classifier processes fault location identification. The simulation results show that the proposed method achieves a high classification rate for considering fault inception angles and fault resistance uncertainty.

On diagnostic accuracy measure with cut-points criterion for ordinal disease classification based on concordance and discordance

The accuracy of a diagnostic test has always been essential in detecting disease staging. Many diagnostic tests of accuracy measures are used in binary diagnosis tests. Some measures apply to multi-stage diagnosis. Yet, there are limitations to the implementation, and the performance highly depends on the distribution of diagnostic outcomes. Another essential aspect of medical diagnostic testing using biomarkers is to find an optimal cut-point that categorizes a patient as diseased or healthy. This aspect was extended to the diseases with more than two stages. We propose a diagnostic accuracy measure and optimal cut-points selection (CD), using concordance and discordance for k-stages diseases. The CD measure uses the classification agreement and disagreement between tests outcomes and diseases stages. Simulations for power studies suggest that CD can detect the differences between the null and alternative hypotheses that other methods cannot for some scenarios. Simulation results indicate that using CD measures to select optimal cut-points can provide relatively high correct classification rates than the existing measures and more balanced accurate classification rates than the generalized Youden Index (GYI). An illustration is provided using the ANDI data to choose biomarkers for diagnosing Alzheimer's Disease (AD) and select optimal cut-points for the chosen biomarkers.

Hologram classification of occluded and deformable objects with speckle noise contamination by deep learning.

Advancements in optical, computing, and electronic technologies have enabled holograms of physical three-dimensional (3D) objects to be captured. The hologram can be displayed with a spatial light modulator to reconstruct a visible image. Although holography is an ideal solution for recording 3D images, a hologram comprises high-frequency fringe patterns that are almost impossible to recognize with traditional computer vision methods. Recently, it has been shown that holograms can be classified with deep learning based on convolution neural networks. However, the method can only achieve a high success classification rate if the image represented in the hologram is without speckle noise and occlusion. Minor occlusion of the image generally leads to a substantial drop in the success rate. This paper proposes a method known as ensemble deep-learning invariant occluded hologram classification to overcome this problem. The proposed new method attains over 95% accuracy in the classification of holograms of partially occluded handwritten numbers contaminated with speckle noise. To achieve the performance, a new augmentation scheme and a new enhanced ensemble structure are necessary. The new augmentation process includes occluded objects and simulates the worst-case scenario of speckle noise.

Multi-Class Brain Disease Classification Using Modified Pre-Trained Convolutional Neural Networks Model with Substantial Data Augmentation

The integration of various algorithms in the medical field to diagnose brain disorders is significant. Generally, Computed Tomography, Magnetic Resonance Imaging techniques have been used to diagnose brain images. Subsequently, segmentation and classification of brain disease remain an exigent task in medical image processing. This paper presents an extended model for brain image classification based on a Modified pre-trained convolutional neural network model with extensive data augmentation. The proposed system has been efficiently trained using the technique of substantial data augmentation in the pre-processing stage. In the first phase, the pre-trained models namely AlexNet, VGGNet-19, and ResNet-50 are employed to classify the brain disease. In the second phase, the idea of integrating the existing pre-trained model with a multiclass linear support vector machine is incorporated. Hence, the SoftMax layer of pre-trained models is replaced with a multi class linear support vector machine classifier is proposed. These proposed modified pre-trained model is employed to classify brain images as normal, inflammatory, degenerative, neoplastic and cerebrovascular diseases. The training loss, mean square error, and classification accuracy have been improved through the concept of Cyclic Learning rate. The appropriateness of transfer learning has been demonstrated by applying three convolutional neural network models, namely, AlexNet, VGGNet-19, and ResNet-50. It has been observed that the modified pre-trained models achieved a higher classification rate of accuracies of 93.45% when compared with a finetuned pre-trained model of 89.65%. The best classification accuracy of 92.11%, 92.83% and 93.45% has been attained in the proposed method of the modified pre-trained model. A comparison of the proposed model with other pre-trained models is also presented.

Multi-information fingerprint identification method based on interactive genetic algorithm

In order to overcome the shortcomings of traditional fingerprint identification methods, such as low computational efficiency and high false classification rate, a new multi-information fingerprint identification method based on interactive genetic algorithm is proposed. Firstly, the multi-information fingerprint image is pre-processed to extract the feature points. Then, combined with the interactive genetic algorithm, the rotation angle of the two fingerprint images is determined. At the same time, the fingerprint offset code is coded, and the matching setting function is set. The matching degree of the two fingerprints is determined by the interactive genetic algorithm, which effectively realises the multi-information fingerprint identification. Finally, the simulation experiment is carried out. The experimental results show that the proposed method can effectively reduce the false classification rate, reduce the average matching time of fingerprint image, and improve the operation efficiency. The minimum error rate is only 1.02%.

A Comparative Study of Machine Learning-based Approach for Network Traffic Classification

Internet usage has increased rapidly and become an essential part of human life, corresponding to the rapid development of network infrastructure in recent years. Thus, protecting users’ confidential information when joining the global network becomes one of the most significant considerations. Even though multiple encryption algorithms and techniques have been applied in different parties, including internet providers, and web hosting, this situation also allows the hacker to attack the network system anonymously. Therefore, the significance of classifying network data streams to improve network system quality and security is attracting increasing study interests. This work introduces a machine learning-based approach to find the most suitable training model for network traffic classification tasks. Data pre-processing is first applied to normalize each feature type in the dataset. Different machine learning techniques, including k-Nearest Neighbors (KNN), Artificial Neural Network (ANN), and Random Forest (RF), are applied based on the normalized features in the classification phase. An open-access dataset ISCXVPN2016 is applied for this research, which includes two types of encryption (VPN and Non-VPN) and seven classes of traffic categories classes. Experimental results on the open dataset have shown that the proposed models have reached a high classification rate – over 85% in some cases, in which the RF model obtains the most refined results among the three techniques.

A Novel Local Feature Generation Technique Based Sound Classification Method for Covid-19 Detection using Lung Breathing Sound

Lung breathing sounds have been used to diagnose many diseases, including Covid-19. Nowadays, Covid-19 has affected daily life worldwide, and it has caused a global pandemic. Generally, computer vision methods have been presented to classify healthy, pneumonia, and Covid-19. They achieved high classification rates on datasets with a limited number of classes without taking into consideration other lung diseases. Our main hypothesis is to detect Covid-19 automatically among other lung diseases by using lung breathing sounds. Therefore, a dataset of lung breathing sound with ten classes has been collected, and a novel lung sounds classification method has been proposed in this paper. This method presents a novel local feature generation technique, and Substitution Box (S-Box) of the present lightweight encryption method is utilized as a pattern. A novel nonlinear pattern is presented based on S-Box, named Present-SBox-Pat (present S-Box pattern). A new pooling-based transformation (maximum tent pooling (MaTP)) is proposed to generate high, middle, and low levels features. It is considered as a preprocessing method of this work. ReliefF and iterative neighbourhood component analysis (RFINCA) selector is used to select the most discriminative and informative features. Two shallow classifiers are used to obtain results. The proposed Present-SBox-Pat and MaTP feature generation network and RFINCA feature selector-based method achieved 95.43% classification accuracy using the SVM classifier. These results demonstrated the success of techniques in generating and selecting features that facilitate the task of classifiers.

Diagnosis of Brain Tumor Using Artificial Neural Network

Accurate detection and diagnosis of brain tumor is one the crucial task of medical image analysis. Brain tumor classification system aids the physician to make accurate diagnosis and to provide effective treatment. Magnetic Resonance Imaging (MRI) is the gold standard imaging technique for brain tumor diagnosis. This paper proposes a method for brain tumor detection and classification using artificial neural network. The proposed method consists of four major processes such as preprocessing, region of interest segmentation, feature extraction and classification. Feed forward neural network is employed to classify the brain tumors. Classification performance of the proposed method is evaluated using 10-cross fold validation and compared with the previous methods. Empirical findings proved that the proposed method can efficiently classify the brain tumor with higher classification rate.

Prediction of neuro-degenerative disorders using sunflower optimisation algorithm and Kernel extreme learning machine: A case-study with Parkinson's and Alzheimer's disease.

Parkinson's and Alzheimer's Disease are believed to be most prevalent and common in older people. Several data-mining approaches are employed on the neuro-degenerative data in predicting the disease. A novel method has been built and developed to diagnose Alzheimer's (AD) and Parkinson's (PD) in early stages, which includes image acquisition, pre-processing, feature extraction and selection, followed by classification. The challenge lies in selecting the optimal feature subset for classification. In this work, the Sunflower Optimisation Algorithm (SFO) is employed to select the optimal feature set, which is then fed to the Kernel Extreme Learning Machine (KELM) for classification. The method is tested on the Alzheimer's Disease Neuroimaging Initiative (ADNI) and local dataset for AD, the University of California, Irvine (UCI) machine learning repository and the Istanbul dataset for PD. Experimental outcomes have demonstrated a high accuracy level in both AD and PD diagnosis. For AD diagnosis, the highest classification rate is obtained for the AD versus NC classification using the ADNI dataset (99.32%) and local dataset (98.65%). For PD diagnosis, the highest accuracy of 99.52% and 99.45% is achieved on the UCI and Istanbul datasets, respectively. To show the robustness of the method, the method is compared with other similar methods of feature selection and classification with 10-fold cross-validation (CV) and with unseen data. The method proposed has an excellent prospect, bringing greater convenience to clinicians in making a better solid decision in clinical diagnosis of neuro-degenerative diseases.

Automatic Detection and Classification of Diabetes Using Artificial Intelligence

Diabetes is characterized as a chronic disease that may cause many health complications. Artificial intelligence techniques are adopted diagnose diabetes more accurately. This paper presents an artificial intelligence technique for diabetes diagnosis. Efficacy of the technique is evaluated using diabetes database. Experimental results show that the back propagation neural network algorithm yields the highest classification rate compared to k-nearest neighbourhood classifier. Additionally, the back propagation neural network provides error with the highest area under curve of 90 %.

Quality of commercial soybean grain and post-harvest economic losses in farming in West Pará, Brazil

The soybean production chain is one of the main agribusiness products. In this context, the grain market is constantly expanding. Thus, it is important to carry out the characterization of soybeans and define their qualities. The objective of this study was to evaluate the quality of commercial soybean grains in post-harvest, as well as the economic losses in a crop in the municipality of Mojuí dos Campos, Pará. Two types of cultivars at 4280(IPRO) and 8644 (IPRO) were used, harvested on different days and plots.Data collection was carried out by monitoring the post-harvest process. The information regarding the grains was obtained through the classification report provided by the company accredited to classify the grains. The methods used were carried out in accordance with the Technical Regulation of Soybeans of Normative Instruction No. 11, of May 15, 2007, of the Ministry of Agriculture, Pecuária e Abastecimento (MAPA). Data were analyzed by PROC TTEST in SAS version 19.0 at a significance level of 5%. The results showed that, in the period of harvesting of cultivar 8644(IPRO), there was a higher rate of rainfall compared to the days of harvesting of cultivar 4280(IPRO). The total of discounted grains was 15.101t for the cultivar 4280(IPRO). The wet grain is the only parameter responsible for the discount in both cultivars. The total of discounted grains referring to cultivar 8644(IPRO) was 30.218t. The wet grains for the cultivar had the highest classification rate on the fourth day of harvest, with a total of 23.20%, being related to the decrease in the percentage of broken grains (1.60%) and a higher index of burnt/damaged grains. It is concluded that the cultivar 4280(IPRO) presented a lower grain discount compared to the cultivar 8644(IPRO). Consequently, cultivar 8644 presented greater economic damage to the property.

Deep learning approaches for force feedback based void defect detection in friction stir welding

The Friction Stir Welding (FSW) process is known as a solid state welding process comparably stable against external disturbances in its steady state. Therefore, the process is commonly applied with fixed welding parameters utilizing axial force control or position control strategies. External and internal process disturbances, introduced by the workpieces, gap tolerances, tool wear or machine/tool inadequacies are rarely monitored and no conclusion towards the weld seam quality is drawn. In most applications the sole process monitoring feature is the axial force and quality control measures are often only applied to external weld seam features. Based on prior analytic process analysis and quality models based on analytic algorithms, machine learning approaches are employed to find suitable quality criteria to detect internal weld seam defects (void defects) over a wide range of welding parameters. The recorded welding force data is examined using Long Short-Term Memory (LSTM) Networks, Bidirectional Long Short-Term Memory (BiLSTM) Networks and different versions of Convolutional Neural Networks (CNN), differing in Layer architecture and Node count (and thus resulting complexity). In an effort to reduce data and speed up development and training, manual feature extraction was performed. Features were selected based on the cyclical nature of the FSW-process and prior research into in-plane process force data quality relevancy. The features as well as raw recorded force data were used to train previously mentioned networks. Overall the viability of the approaches was shown by a high rate of correct void defect classification, with CNNs exceeding 93% test set classification rate. Network training had to be closely monitored as a small dataset and large number of input features increased the risk of overfitting.

Size Distribution, Length-Weight Relationship, and Size at Morphometric Maturity of the Mud Spiny Lobster Panulirus polyphagus (Herbst, 1793) in the Johor Strait

Spiny lobsters of the genusPanulirusare economically important and support local fishing communities. However, mud spiny lobsterPanulirus polyphagusis among the least known species within this genus in terms of their biological information. This study relates to the size distribution, length-weight relationship, and size at morphometric maturity ofP. polyphagusin the Johor Strait. Within the year 2010, 300 specimens were collected off the coast of Johor Strait, Malaysia. There was no significant difference in body size (cephalothorax length, CL) and body weight (BW) between sexes. CL and BW were highly correlated and males ofP. polyphagusdisplayed positive growth allometry, whereas the opposite was observed in females. Based on the merus (ML) and carpus length (CPL) of the third right walking leg, the piecewise linear regression analysis estimated that the size at maturity for male was 6.58 cm CL (based on ML) and 7.58 cm CL (based on CPL), whereas it was 8.18 cm CL (based on ML) and 6.75 cm (based on CPL) for females. Two discriminant functions of high classification and revalidation rates (> 98.6% in males and > 98.7% in females) that can discern maturation status in males and females ofP. polyphaguswere derived using the discriminant function analysis. Biological information derived from this study serves as an essential baseline for future fishery management and conservation ofP. polyphagus.

Texture classification using Minkowski distance measure-based clustering for feature selection

Texture classification has been one of the problem-solving strategies. The feature selection technique is critical for increasing the efficiency of machine learning because it significantly improves the performance of texture classification by removing irrelevant features from the original collection. The majority of methods for selecting features are statistical in nature. Accuracy in database classification is primarily achieved by selecting attributes and simultaneously increasing the classification rate. The task’s primary objective is to select the most significant features from the feature database for execution. We use the cluster-based feature selection method to derive features for statistical and structural approaches. The proposed strategy should be implemented in three stages. To begin, the function can be extracted using the local binary pattern (LBP), the gray level co-occurrence matrix, or the Gabor filter. Following that, a modified K-means clustering method is used to select features using four standard distance measurements: Euclidean, Minkowski, Chebychev, and City Block. Finally, benchmark classifiers such as naive bayes (NB), support vector machine (SVM), and K-nearest neighbor (KNN) were used to effectively compute the classification accuracy of the research. The proposed approach achieves accuracy of 99.4328 for KNN, 98.6654 for NB, and 99.6712 for SVM with the highest classification rate.

Refined Color Texture Classification Using CNN and Local Binary Pattern

Representation and classification of color texture generate considerable interest within the field of computer vision. Texture classification is a difficult task that assigns unlabeled images or textures to the correct labeled class. Some key factors such as scaling and viewpoint variations and illumination changes make this task challenging. In this paper, we present a new feature extraction technique for color texture classification and recognition. The presented approach aggregates the features extracted from local binary patterns (LBP) and convolution neural network (CNN) to provide discriminatory information, leading to better texture classification results. Almost all of the CNN model cases classify images based on global features that describe the image as a whole to generalize the entire object. LBP classifies images based on local features that describe the image’s key points (image patches). Our analysis shows that using LBP improves the classification task when compared to using CNN only. We test the proposed approach experimentally over three challenging color image datasets (ALOT, CBT, and Outex). The results demonstrated that our approach improved up to 25% in the classification accuracy over the traditional CNN models. We identify optimal combinations for each dataset and obtain high classification rates. The proposed approach is robust, stable, and discriminatory among the three datasets and has shown enhancement in classification and recognition compared to the state-of-the-art method.

Intra- and Inter-Specific Variation in Edible Jellyfish Biomarkers and Implications for Origin Traceability and Authentication

With the continuous development of jellyfish fisheries and food products around the world, an effective traceability system has become increasingly prominent. This study provides insight into the origin traceability and authentication of two commercially important jellyfish species, flame jellyfish Rhopilema esculentum and Nomura’s jellyfish Nemopilema nomurai, while investigating the intra- and inter-specific variation in fatty acid (FA) profiles and carbon and nitrogen stable isotope ratios (δ13C and δ15N). Results showed significant differences in FA profiles and isotopic values in fresh bell tissues between wild and farmed R. esculentum and among geographic origins, possibly due to different food sources, nutritional status, and energy costs that each group experiences at a given location. The linear discriminant analysis indicated that δ13C, δ15N, C16:0, C17:0, C18:0, C16:1n7, and C20:5n3 were suitable discriminatory variables with a high rate of correct classification for distinguishing origins of R. esculentum. In addition, inter-specific FA profiles/biomarkers, combined with isotopic values, suggests the variety of dietary sources and trophic positions of sympatric similar-sized R. esculentum and N. nomurai and the potential use of biomarkers, especially stable isotope analysis, for distinguishing sympatric jellyfish species. These results highlighted the complementarity of FA and stable isotope analyses and provide an alternative approach for improving the origin traceability and authenticity evaluation of untreated edible jellyfish. Furthermore, this study adds new information regarding the biochemical compositions of jellyfish species.

Modular deep neural networks for automatic quality control of retinal optical coherence tomography scans

Retinal optical coherence tomography (OCT) with intraretinal layer segmentation is increasingly used not only in ophthalmology but also for neurological diseases such as multiple sclerosis (MS). Signal quality influences segmentation results, and high-quality OCT images are needed for accurate segmentation and quantification of subtle intraretinal layer changes. Among others, OCT image quality depends on the ability to focus, patient compliance and operator skills. Current criteria for OCT quality define acceptable image quality, but depend on manual rating by experienced graders and are time consuming and subjective. In this paper, we propose and validate a standardized, grader-independent, real-time feedback system for automatic quality assessment of retinal OCT images. We defined image quality criteria for scan centering, signal quality and image completeness based on published quality criteria and typical artifacts identified by experienced graders when inspecting OCT images. We then trained modular neural networks on OCT data with manual quality grading to analyze image quality features. Quality analysis by a combination of these trained networks generates a comprehensive quality report containing quantitative results. We validated the approach against quality assessment according to the OSCAR-IB criteria by an experienced grader. Here, 100 OCT files with volume, circular and radial scans, centered on optic nerve head and macula, were analyzed and classified. A specificity of 0.96, a sensitivity of 0.97 and an accuracy of 0.97 as well as a Matthews correlation coefficient of 0.93 indicate a high rate of correct classification. Our method shows promising results in comparison to manual OCT grading and may be useful for real-time image quality analysis or analysis of large data sets, supporting standardized application of image quality criteria.

Pap smear based cervical cancer detection using residual neural networks deep learning architecture

AbstractDetecting and classifying the Pap smear cell images is important task for cervical cancer identification. In this article, dual tree complex wavelet transform (DTCWT) based modified deep learning algorithm is proposed for classifying the Pap smear cell images into four different cases normal, carcinoma in situ, dysplastic, and superficial. This proposed work consists of data augmentation module, DTCWT module and convolutional neural networks (CNN) module for the automatic classification of Pap smear cell images. The CNN module required huge number of cell images for obtaining high classification rate. Hence, the shearing and flipping functions are used in the data augmentation process to improve the dataset samples for further CNN classification module. Then, the spatial pixel behavior of data augmented images is transferred into multimodal pixels using DTCWT, which creates the sub band coefficients in matrix format. This matrix is trained and classified using ResNet 18 which classifies the source Pap smear cell image into four classes. This developed method is tested on the Pap smear cell images available in open access dataset. The average Pap smear detection index (PDI) of the automatic Pap smear cell image classification system is about 99%.