Classification Method Research Articles

The classification of melanocytic gene signatures.

Gene expression profiling technologies have revolutionized cell biology, enabling researchers to identify gene signatures linked to various biological attributes of melanomas, such as pigmentation status, differentiation state, proliferative versus invasive capacity, and disease progression. Although the discovery of gene signatures has significantly enhanced our understanding of melanocytic phenotypes, reconciling the numerous signatures reported across independent studies and different profiling platforms remains a challenge. Current methods for classifying melanocytic gene signatures depend on exact gene overlap and comparison with unstandardized baseline transcriptomes. In this study, we aimed to categorize published gene signatures into clusters based on their similar patterns of expression across clinical cutaneous melanoma specimens. We analyzed nearly 800 melanoma samples from six gene expression repositories and developed a classification framework for gene signatures that is resilient against biases in gene identification across profiling platforms and inconsistencies in baseline standards. Using 39 frequently cited published gene signatures, our analysis revealed seven principal classes of gene signatures that correlate with previously identified phenotypes: Differentiated, Mitotic/MYC, AXL, Amelanotic, Neuro, Hypometabolic, and Invasive. Each class is consistent with the phenotypes that the constituent gene signatures represent, and our classification method does not rely on overlapping genes between signatures. To facilitate broader application, we created WIMMS (what is my melanocytic signature, available at https://wimms.tanlab.org/), a user-friendly web application. WIMMS allows users to categorize any gene signature, determining its relationship to predominantly cited signatures and its representation within the seven principal classes.

QCL Infrared Spectroscopy Combined with Machine Learning as a Useful Tool for Classifying Acetaminophen Tablets by Brand

The development of new methods of identification of active pharmaceutical ingredients (API) is a subject of paramount importance for research centers, the pharmaceutical industry, and law enforcement agencies. Here, a system for identifying and classifying pharmaceutical tablets containing acetaminophen (AAP) by brand has been developed. In total, 15 tablets of 11 brands for a total of 165 samples were analyzed. Mid-infrared vibrational spectroscopy with multivariate analysis was employed. Quantum cascade lasers (QCLs) were used as mid-infrared sources. IR spectra in the spectral range 980–1600 cm−1 were recorded. Five different classification methods were used. First, a spectral search through correlation indices. Second, machine learning algorithms such as principal component analysis (PCA), support vector classification (SVC), decision tree classifier (DTC), and artificial neural network (ANN) were employed to classify tablets by brands. SNV and first derivative were used as preprocessing to improve the spectral information. Precision, recall, specificity, F1-score, and accuracy were used as criteria to evaluate the best SVC, DEE, and ANN classification models obtained. The IR spectra of the tablets show characteristic vibrational signals of AAP and other APIs present. Spectral classification by spectral search and PCA showed limitations in differentiating between brands, particularly for tablets containing AAP as the only API. Machine learning models, specifically SVC, achieved high accuracy in classifying AAP tablets according to their brand, even for brands containing only AAP.

Editorial

Dear Readers, It gives me great pleasure to announce the seventh regular issue of 2024. In this issue, 5 papers by 12 authors from 6 countries - Brazil, Ecuador, Germany, Iraq, Spain, Turkey - cover various topical aspects of computer science. In a continuous effort to further strengthen our journal, I would like to expand the editorial board: If you are a tenured associate professor or above with a strong publication record, you are welcome to apply to join our editorial board. We are also interested in high-quality proposals for special issues on new topics and trends. As always, I would like to thank all the authors for their sound research and the editorial board and guest reviewers for their extremely valuable review effort and suggestions for improvement. I also want to thank the readers for their interest in our articles, which is reflected in the consistently high number of user accesses and PDF downloads. These contributions, together with the generous support of the consortium members, maintain the quality of our journal. In the seventh regular issue, I am very pleased to present the following 5 accepted articles: Vinicius Bischoff and Kleinner Farias from Brazil report their study on a controlled experiment with 22 participants (15 students, 7 professionals) in which the correctness and the effort required for the integration of feature models were investigated. Abdulkadir Buldu, Kaplan Kaplan and Melih Kuncan from Turkey present their research on an assistive system based on EEG data and Continuous Wavelet Transform (CWT), which aims to reduce the life-threatening risk for epilepsy patients. In a collaboration between researchers from Ecuador and Spain, Darwin Alulema, Maximiliano Paredes-Velasco and Ricardo de Arriba Lasso report in their manuscript on the LESCA system, which performs adaptive content feedback through scaffolding and supports the development of high-level competencies. In another joint research between colleagues from Iraq and Germany, Anwar Mira and Olaf Hellwich present their approach to improve recognition capabilities by optimizing deep learning features for hand gesture image recognition. Specifically, they propose to enhance features of well-trained DNNs using an improved radial basis function (RBF) neural network, targeting recognition within individual gesture categories. And last but not least, Rasim Çekik and Mahmut Kaya from Turkey propose in their research a new performance metric to evaluate filter feature selection methods in text classification.  Enjoy Reading! Best regards,  Christian Gütl, Managing Editor 

A New Performance Metric to Evaluate Filter Feature Selection Methods in Text Classification

High dimensionality and sparsity are the primary issues in text classification. Using feature selection approaches, the most effective way to solve the problem is to select a subset of features. The most common and effective methods used for this process are filter techniques. Various performance metrics such as Micro-F1, Macro-F1, and Accuracy are used to evaluate the performance of filter methods used for feature selection on datasets  Such methods work depending on a classification algorithm. However, when selecting features in filter techniques, the information on the individual features is evaluated without considering the relationship between the features. In such an approach, the actual performance of the filter technique used in feature selection may not be determined. In such a case, it causes the existing methods to be insufficient in testing the validity of the proposed method. For this purpose, this study suggests a novel performance metric called Selection Error (SE) to determine the actual performance evaluation of filter techniques. The Selection Error metric allows us to analyze the information value of the selected features more accurately than existing methods without relying on a classifier. The feature selection performance of the filtering approaches was performed on six different datasets with both The Selection Error and traditional performance metrics. When the results are examined, it is seen that there is a strong relationship between the proposed performance metric and the classification performance metric results. The Selection Error aims to significantly contribute to the literature by demonstrating the success of filtering feature selection methods, regardless of classifier performance. 

An Improve Method for Plant Leaf Disease Detection and Classification using Deep Learning

In countries like India, whose important occupation is agriculture, face a huge loss when the crops get affected by any type of disease. These diseases attack the crops in various stages and can destroy the entire production. Since most diseases are transmitted from one crop to another there is an essential requirement to detect the type of disease in the early stage so that farmers can take the required action to “save the crops” and production. Early disease detection is one of the essential activities for enhancing agricultural productivity. Diseases spread very quickly in the parts of the leaves that affect the growth of the plants. Early detection is a challenging task as the symptoms are mild for accurate identification. This research paper presents an enhanced CNN based MCC-ECNN model with fine-tuned hyper-parameters and various batch sizes for accurate plant leaf disease classification.

LW-XNet for segmentation and classification of skin lesions from dermoscopy images

The skin, one of the most crucial organs of the human body, serves as a barrier between the body and the external environment. Early detection of skin diseases is imperative to reduce mortality rates, as some untreated conditions may progress to skin cancer. Segmentation and classification of lesions represent pivotal and interrelated endeavors within the realm of skin disease diagnosis. For this purpose, this paper presents a comprehensive diagnostic framework for segmentation and classification of skin lesions, which integrates a Legendre multiwavelet transform-based fusion XNet (LW-XNet) with an improved soft attention dense connection convolutional network (ISA-DenseNet). LW-XNet combines the strengths of XNet in fusing different frequency components of images, along with the strong feature representation capability of LW bases with various regularities for overall contextual information and detailed information of dermoscopy images. Furthermore, its encoder devises a LWT channel concatenate (LCC) block to subdivide the image into eight wavelet coefficient feature images and perform concatenated processing on them, enabling it to better differentiate and comprehend the intricate features within dermoscopy images. Finally, ISA-DenseNet is utilized for multi-class classification of the segmented images. Experimental results demonstrate the superiority of the proposed framework over existing segmentation and classification methods.

LULC change detection analysis of Chamarajanagar district, Karnataka state, India using CNN-based deep learning method

The change detection analysis of land use land cover (LULC) is an important task in several fields and applications such as environmental monitoring, urban planning, disaster management, and climate change studies. This study focuses on the use of remote sensing (RS) and geographic information systems (GIS) to identify the changes in Chamarajanagar district, which is located in Karnataka state, South India. This paper mainly focuses on the classification and change detection analysis of LULC in 2011 and 2021 using linear imaging self-scanning sensor-III (LISS-III) satellite images. Traditional methods for LULC classification involve manual interpretation of satellite images, which provides lower accuracy. Therefore, this paper proposed the Convolutional Neural Network (CNN)-based deep learning classification method for LULC classification. The main objective of the research work is to perform an accurate change detection of the Chamarajanagar district using the classified maps of the years 2011 and 2021. The proposed classification method is outperformed, with a classification accuracy of 95.27 % and 94.57 % for LISS-III satellite imagery of the years 2011 and 2021 respectively. Further, change detection analysis has been carried out using classified maps and results show a decline of 3.23 sq. km, 22.7 sq. km, and 3.83 sq. km in the areas covered by vegetation, agricultural land, and forest area, respectively. In other classes, such as built-up, water bodies, and barren land, an increase in land cover was observed by 5.59 sq. km, 1.99 sq. km, and 20.92 sq. km, respectively.

Analysis of Product Sales in the Application of Data Mining with Naive Bayes Classification

H&F Shoe Store is a privately owned Micro, Small, and Medium Enterprises retail store that sells merchandise. The owner serves customers directly and also acts as a cashier. In this store, the business owner is less aware of what types or categories of products are most in demand by customers, making sales operations less than optimal. Because of this, special expertise is needed to handle the problems in the retail store, namely data mining or Data Mining with the aim of digging up information related to sales problems, in this case the author will use the Classification method with the Naive Bayes algorithm. In this study, the author uses secondary data obtained from sales notebooks and re-collected into Microsoft Excel according to research needs. The data that has been collected on the software is 121 data which have 10 attributes, namely “Nama Produk”, “Size Produk”, “Kategori Produk”, “Jenis Produk”, “Gender Produk”, “Merek Produk”, “Stok Awal”, “Stok Terjual”, “Stok Sisa”, and “Penjualan”. The Naive Bayes Classifier method has successfully produced good results in classifying sales on a type or category of marketed products, the results obtained are in the form of product sales analysis and Naive Bayes model evaluation values. The results of the model evaluation values on the Confusion Matrix obtained are accuracy of 86.11%, recall of 84.62% and precision of 84.62%.

Reconstruction method for a three-demensional discrete element numerical model of landslides using an integrated multi-electrode resistivity tomography method and an unmanned aerial vehicle survey

To analyze the hazard-causing modes of landslides, this paper proposes a three-dimensional discrete element model reconstruction method that employs an unmanned aerial vehicle survey and multi-electrode resistivity tomography method. To convert the resistivity profile into a material profile, we adopt the peak of the probability density method for material classification and utilize the Haar wavelet transform for image denoising. Subsequently, inverse distance weighting interpolation and the curtain-point method are used to transform two-dimensional profiles into a 3D visualization model. Similarly, the triangular mesh boundary can be extracted from the 3D visualization model using the curtain-point method. A mapping function f including the macroscopic parameters, was defined to populate the particles within the boundaries. Using the iterative method and defining the loss function L for parameter calibration, the targeted 3D discrete element model was constructed after setting the velocity threshold. This method was applied to the Changhe landslide (September 14, 2019) in Gansu Province, China, which had a typical damaged soil layer due to earthquake and rainfall factors. The results indicate that the lower part first exhibits significant displacement, followed by the upper and middle parts, which is consistent with the on-site inspections and UAV findings.

MM_Fast_RCNN_ResNet: Construction of Multimodal Faster RCNN Inception and ResNet V2 for Pedestrian Tracking and detection

Pedestrian identification and tracking is a crucial duty in smart building monitoring. The development of sensors has led to architects' focus on smart building design. The image distortions caused by numerous external environmental factors present a significant problem for pedestrian recognition in smart buildings. It is difficult for machine learning algorithms and other conventional filter-based image classification methods, such as histograms of oriented gradient filters, to function efficiently when dealing with many input photos of pedestrians. Deep learning algorithms are now performing substantially better when processing an enormous amount of image data. This article evaluates a novel multimodal classifier-based pedestrian identification method. The proposed method is Multimodal Faster RCNN Inception and ResNet V2 (MM Fast RCNN ResNet). The collected attributes address a tracking problem and establish the foundation for several object recognition tasks (novelty). Our method's neural network is regularized, and the feature representation is automatically adjusted to the detection assignment, resulting in high accuracy (superior to the proposed method). The proposed method is assessed using the PenFudan dataset and contemporary techniques regarding several factors. It is discovered that the recommended MM Fast RCNN ResNet obtains precision, recall, FPPI, FPPW, and average precision of 0.9057, 0.8629, 0.0898, and 0.0943.

Study on the fluidity of the pore-fracture binary system in a tight sandstone reservoir-NMR

Fluid movability in tight sands may not be accurately characterized by pore size-based classification methods solely because of the complex pore structure and heterogeneity in pore size. In this study, on the basis of casting thin slices and scanning electron microscope observation, pore structure was analyzed using mercury injection, NMR, and micron CT to classify and evaluate the tight oil reservoir. The experiment suggest that the quality of tight reservoir is determined by its pore structure, particularly the throat radius, with the microthroat being an essential factor in permeability. Uniquely, we divide the reservoir by Q-cluster with throat radius, displacement pressure, permeability and other parameters. Based on reservoir classification, this study proposed a method for studying the pore size classification of samples on the T2 spectrum by combining CT scanning with mercury intrusion and a NMR experiment. Pore fluids are generally classified into movable fluid and irreducible fluid by one or two NMR T2 cut-offs. The pore size distributions and capillarity boundaries are converted from T2 and mercury injection capillary pressure (MICP). We categorized pores into micropores (T2 < 1), macropores (T2 > 10, with T2 > 300 as fractures), and medium pores (the rest). The saturation of movable fluid and the percentage of micro-fractures can characterize the seepage characteristics of tight reservoirs, which is of great significance for the later periods of oilfield development.

Computer Vision Algorithm Design in Image Processing Based on Projective Geometry

Image processing with computer vision, particularly in the realm of projective geometry, offers remarkable potential for various applications. Through the lens of projective geometry, images can be transformed, augmented, and reconstructed with precision, facilitating tasks such as image rectification, 3D reconstruction, and object tracking. Landmark estimation in computer vision is a vital task with broad applications across various domains. This process involves identifying key points or landmarks within images, enabling tasks such as facial recognition, object tracking, and gesture recognition. This paper, proposed a novel approach for landmark estimation in computer vision using Projective Geometry Landmark Estimation (PGLM). The proposed model aims to estimate the landmark features by a projective geometry model. With the estimation of the geometry features landmarks related to the facial, object, and medical images are computed. The PGLM model uses the point features for the location of the landmark features. In order to compare PGLM's performance to that of more conventional classification methods like Random Forest, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM), simulation analysis is carried out. From what we can see, PGLM routinely beats these alternatives when we compare their accuracy, precision, recall, and F1 score. The findings stated the effectiveness of PGLM as a promising approach for landmark estimation in image processing tasks, paving the way for further advancements in this domain.

K-Nearest neighbor smart contract classification with semantic feature enhancement

Abstract How to quickly and accurately retrieve relevant smart contracts from a huge amount of smart contracts has become an urgent need for users. The classification of smart contracts offers a solution by narrowing down the search space. Existing smart contract classification methods suffer from incomplete semantic feature extraction and a lack of consideration of the existence of rich semantics in existing smart contracts of the same class. To address the above problems, we propose a contrast learning and semantic feature embedding approach to enhance K-Nearest Neighbor (CL-SFE-IKNN). Our method fuses local features, global features, and account transaction features of the smart contract source code to perfect the semantics of the contract. Our method adopts KNN to retrieve multiple instances of contracts in the same class and assigns weights to the model output based on their labels. Meanwhile, we introduce contrastive learning and semantic feature embedding to enhance KNN retrieval to high-quality nearest neighbors of the same class. Experimental results show that by combining a KNN classifier with a traditional linear classifier, our model achieves the best performance compared with other baseline models.

A new dynamic classifier selection method for text classification

A new dynamic classifier selection method for text classification

The main aspects of black hole theory

Black holes are extremely compact objects which no one knows exactly what it is, and astrophysicists, whether in the past or the present, have been devoting a great deal of time to studying them. The establishment of black hole theory takes a lot of time. There are quite some difficulties in breaking continuous negation, admitting, and researching black holes. In this paper, we first look back at the discovery process of black hole theory. We mainly review the three aspects of black holes: physical properties, classification and observational methods. Black holes are invisible and many astrophysicists in the past only can investigate them based on theoretical understanding, so we present some new and clear evidences of black holes to prove this theory. In this paper, some new models and applications of black holes (e. g. Hawking Radiation Theory, The Penrose Process) that may be discovered in the future will be presented too.

Code smells analysis for android applications and a solution for less battery consumption

In the digitization era, the battery consumption factor plays a vital role for the devices that operate Android software, expecting them to deliver high performance and good maintainability.The study aims to analyze the Android-specific code smells, their impact on battery consumption, and the formulation of a mathematical model concerning static code metrics hampered by the code smells. We studied the impact on battery consumption by three Android-specific code smells, namely: No Low Memory Resolver (NLMR), Slow Loop (SL) and Unclosed Closable, considering 4,165 classes of 16 Android applications. We used a rule-based classification method that aids the refactoring ideology. Subsequently, multi-linear regression (MLR) modeling is used to evaluate battery usage against the software metrics of smelly code instances. Moreover, it was possible to devise a correlation for the software metric influenced by battery consumption and rule-based classifiers. The outcome confirms that the refactoring of the considered code smells minimizes the battery consumption levels. The refactoring method accounts for an accuracy of 87.47% cumulatively. The applied MLR model has an R-square value of 0.76 for NLMR and 0.668 for SL, respectively. This study can guide the developers towards a complete package for the focused development life cycle of Android code, helping them minimize smartphone battery consumption and use the saved battery lives for other operations, contributing to the green energy revolution in mobile devices.

Designing and Prototyping of AI-based Real-time Mobile Detectors for Calisthenic Push-up Exercise

Fitness exercises, including push-ups, are very beneficial to personal health. Many Artificial Intelligence (AI)-based fitness trainers are developed based on human pose estimation models or assisted by Internet of Things (IoT) devices. However, many of them require access to a graphing processing unit (GPU) for model training or IoT sensors to deploy, less accessible for individuals. In our work, we designed and prototyped real-time mobile push-up detectors using three distinctive approaches: (1) Push-up pose classification, (2) Angle-heuristic estimation and (3) Optical flow detection. We trained our deep-learning model with over 2000 images to achieve a high accuracy for real time deployment. Models are tested on our video dataset applied data augmentation techniques to simulate real-world environmental conditions to evaluate model performance based on accuracy metrics (precision, recall, F1 score) and processing frame rate (FPS). From the results, we concluded that the angle-heuristic estimation method has the best overall performance and we analysed the reasons for the relatively poorer performance of the push-up pose classification and optical flow detection methods. All methods developed are capable of working on mobile devices without the need of GPU or IoT sensors.

Key updates and interpretation of the ninth version of AJCC staging system for neuroendocrine tumors of the stomach

The cancer staging system of the American Joint Committee on Cancer (AJCC) is the most widely used clinical basis for tumor staging. In October 2023, AJCC released the staging system (ninth version) for the neuroendocrine tumors of stomach (NET), which has been implemented in January 2024. The ninth version of NET staging system mainly updated the histopathologic classification, diagnosis and staging methods, clinical and pathological staging, prognosis grade, tumor and non-tumor prognostic features. The update and implementation of the staging system provide a more detailed reference for the accurate diagnosis, staging and precise treatment of gastric neuroendocrine tumors. Moreover, it is convenient for clinicians to carry out clinical practice. The purpose of our article is to provide a high-level overview of the major changes in AJCC staging system (version 9) for gastric NET based on the latest evidence-based medical research.

A proposal for cut marks classification using machine learning: Serrated vs. non-serrated, single vs. double-beveled knives.

In tool mark identification, there is still a lack of characteristics and methodologies standardization used to analyze and describe sharp force trauma marks on skeletal remains. This study presents a classification method for cut marks on human bones, providing an applicable methodology for their examination and the relevant terminology for describing cases of sharp force trauma. A total of 350 cut marks were produced by stabbing pig ribs (Sus scrofa) with seven knives. The samples were analyzed under a stereomicroscope with a tangential light source. Through the analysis of cut marks, eleven traits were identified as significantly associated with the type of knife used. These traits included the general morphology of the kerf shape, the entrance and exit cross-profile shapes, the location of the rising on the entrance and exit cross-profile, the presence or absence of feathering, the presence or absence of shards and the location and the general morphology of the mounding. Binary logistic regression models were later trained and tested using nine out of the eleven traits. The first model categorized the cut mark as either produced by a serrated or non-serrated blade, while the second, as either produced by a single- or double-beveled blade. Classification scores of those models ranged between 63%-85% for the serration class and 63%-89% for the blade bevel class. This study proposes a new set of traits and the use of machine learning models to standardize and facilitate the analysis of stab wounds.

Artificial Intelligence in primary care: Intelligent risk predicting platform for non-communicable chronic diseases

Chronic Non-Communicable Diseases with multifactorial and non-infectious characteristics in their origin have been a matter of growing concern for society and governments, as they put people at greater risk of complications, disability, and death. They are the leading cause of death worldwide, i.e. a global threat to health. Modern medicine relies heavily on a variety of biochemical, behavioral, clinical, etc. data to advance the chronic diseases field. Machine learning techniques have a wide range of applications, for example for diabetes classifications, phenotypic resistance of whole genomic sequences and heart disease. These approaches, in general, concentrate efforts on a single Chronic Non-Communicable Diseases. Thus, identifying multiple chronic diseases simultaneously is an open challeng. Our study demonstrated that the intelligent risk prediction platform for non-communicable chronic diseases using Multi-Label Classification methods can be used to model multiple Chronic Non-Communicable Diseases. We use ensemble methods developed on top of algorithm adaptation techniques, including Random Forest, Extra Trees and Decision Tree Classifiers. Among the models in the experiment, the Random Forest achieved the best accuracy and F1-score performance with 96.16% and 90.48%, respectively. Our results suggest that the platform is promising and can be used to support medical teams in the early diagnosis of Chronic Non-Communicable Diseases, allowing them to provide adequate treatment options to patients in a timely manner in primary care, as well as contributing to reducing the risk of complications, disability, and deaths by Chronic Non-Communicable Diseases.