Classification Process Research Articles

Adjudication of Self-reported Symptomatic Stone Recurrence in the Prevention of Urinary Stones With Hydration Trial

ObjectiveTo assess accuracy of self-reported stone events in a large clinical trial by adjudication against the weight of documentation for spontaneous stone passage or surgical intervention. MethodsParticipants in the Prevention of Urinary Stones with Hydration (PUSH) trial were randomized to a multi-component behavioral intervention or control arm to increase and maintain high fluid intake. The primary endpoint was urinary stone events including symptomatic stone passage or procedural intervention. An independent adjudication committee blinded to randomization assignments reviewed all events. Confirmed clinical stone events required typical stone symptoms and documentation of stone passage (e.g., via photograph, clinical record) and/or surgical intervention. Events with typical symptoms and self-described stone passage but without objective documentation of passage were also considered as meeting the primary endpoint and classified separately as patient-reported passage. Non-events did not meet either criteria. ResultsAt time of this blinded analysis, a total of 1658 participants were randomized and had a median follow-up of 19 months. Self-reported stone events (n=217) were adjudicated by the committee as either confirmed clinical events (134; 61.8%), patient-reported passage (71; 32.7%), or non-events (12; 5.5%). Confirmed clinical events consisted of stone passage in 66/134 and procedural interventions in 68/134 (53 for symptoms and 15 without symptoms). ConclusionsRigorous adjudication revealed that self-reported stone events in the PUSH trial overwhelmingly represented clinically documented passage, surgical intervention, and patient-reported passage outside healthcare settings, with only 5.5% failing to satisfy adjudication criteria. Similar adjudication and classification processes warrant consideration for implementation in future stone trials. Clinical Trials RegistrationNCT03244189

Clustered Automated Machine Learning (CAML) model for clinical coding multi-label classification

AbstractClinical coding is a time-consuming task that involves manually identifying and classifying patients’ diseases. This task becomes even more challenging when classifying across multiple diagnoses and performing multi-label classification. Automated Machine Learning (AutoML) techniques can improve this classification process. However, no previous study has developed an AutoML-based approach for multi-label clinical coding. To address this gap, a novel approach, called Clustered Automated Machine Learning (CAML), is introduced in this paper. CAML utilizes the AutoML library Auto-Sklearn and cTAKES feature extraction method. CAML clusters binary diagnosis labels using Hamming distance and employs the AutoML library to select the best algorithm for each cluster. The effectiveness of CAML is evaluated by comparing its performance with that of the Auto-Sklearn model on five different datasets from the Medical Information Mart for Intensive Care (MIMIC III) database of reports. These datasets vary in size, label set, and related diseases. The results demonstrate that CAML outperforms Auto-Sklearn in terms of Micro F1-score and Weighted F1-score, with an overall improvement ratio of 35.15% and 40.56%, respectively. The CAML approach offers the potential to improve healthcare quality by facilitating more accurate diagnoses and treatment decisions, ultimately enhancing patient outcomes.

Fractional differentiation based image enhancement for automatic detection of malignant melanoma

Recent improvements in artificial intelligence and computer vision make it possible to automatically detect abnormalities in medical images. Skin lesions are one broad class of them. There are types of lesions that cause skin cancer, again with several types. Melanoma is one of the deadliest types of skin cancer. Its early diagnosis is at utmost importance. The treatments are greatly aided with artificial intelligence by the quick and precise diagnosis of these conditions. The identification and delineation of boundaries inside skin lesions have shown promise when using the basic image processing approaches for edge detection. Further enhancements regarding edge detections are possible. In this paper, the use of fractional differentiation for improved edge detection is explored on the application of skin lesion detection. A framework based on fractional differential filters for edge detection in skin lesion images is proposed that can improve automatic detection rate of malignant melanoma. The derived images are used to enhance the input images. Obtained images then undergo a classification process based on deep learning. A well-studied dataset of HAM10000 is used in the experiments. The system achieves 81.04% accuracy with EfficientNet model using the proposed fractional derivative based enhancements whereas accuracies are around 77.94% when using original images. In almost all the experiments, the enhanced images improved the accuracy. The results show that the proposed method improves the recognition performance.

Enhancing stroke prediction using the waikato environment for knowledge analysis

<p>State-of-the-art data mining tools incorporate advanced machine learning (ML) and artificial intelligence (AI) models, and it is widely used in classification, association rules, clustering, prediction, and sequential models. Data mining is important for the process of diagnosing and predicting diseases in the early stages, and this contributes greatly to the development of the health services sector. This study utilized classification to predict the stroke of a sample of the patient dataset that was taken from Kaggle. The classification model was created using the data mining program waikato environment for knowledge analysis (WEKA). This data mining tool helped identify individuals most at risk of stroke based on analysis of features extracted from the patient’s dataset. These features were used in classification processes according to the naive Bayes (NB), random forest (RF), support vector machine (SVM), and multi-layer perceptron (MLP) algorithms. Analysis of the classification results of the previous algorithms showed that the SVM outperformed other algorithms in terms of accuracy (94.4%), sensitivity (100%), and F-measure (97.1%). However, the NB algorithm had the best performance in terms of precision (95.7%).</p>

Crime Activity Detection in Surveillance Videos Based on Developed Deep Learning Approach

In modern communities, lots of offenders are prone to recidivism, hence, there is a requirement to inhibit such criminals, especially from impending socioeconomically disadvantaged and high-crime areas that experience elevated levels of criminal activity, involving drug-related offenses, violence, theft, and other forms of anti-social behavior. Consequently, surveillance cameras have been installed in relevant institutions, and further personnel have been provided to monitor videos using various surveillance apparatus. However, relying solely on monitoring with the naked eye and manual video processing falls short of accurately evaluating the footage acquired via such cameras. To handle the issues of conventional systems, there is a need for a system that is able to classify acquired images while supporting surveillance personnel actively. Therefore, in this paper, a deep-learning approach is developed to build a crime detection system. This developed approach includes various layers necessary to perform feature extraction and classification processes and make the system capable of efficiently and accurately detecting crime activities from surveillance video frames. Besides the proposed crime activity detection system, two deep-learning approaches (EfficientNet-B7, and MobileNet-V2) are trained and assessed on the popular UCF Crime and DCSASS datasets. Generally, the proposed detection system encompasses dataset preparation and pre-processing, splitting the pre-processed crime activity image dataset, and implementing the proposed deep learning approach and other pre-trained approaches.

The color features and k-nearest neighbor algorithm for classifying betel leaf image

Piper betle L. (betel) is a species that belongs to the genus Piper and is a type of medicinal plant that is quite well known to the general population. The varieties of the leaf color may distinguish are red, green, and black betel. However, consumers still need assistance determining the differences between the many types of betel leaf. Therefore, using image processing techniques, this research contributes to building a classification method for distinguishing betel leaves based on color attributes. This approach anoints for the region of interest detection, feature extraction, and classification. In addition, three different classifiers, naïve Bayes, support vector machine, and k-nearest neighbors (k-NN), were used during the classification process. The evaluation for this study used a percentage split to divide a total of 180 images between the training and testing phases. The method’s performance provided the highest accuracy value possible, 100%, by utilizing the color characteristics with the k-NN classifier.

A proactive grey wolf optimization for improving bioinformatic systems with high dimensional data

This paper introduces a new methodology for optimization problems, combining the Grey Wolf Optimizer (GWO) with Simi-stochastic search processes. Intelligent optimizations represent an advanced approach in machine learning and computer applications, aiming to reduce the number of features used in the classification process. Optimizing bioinformatics datasets is crucial for information systems that classify data for intelligent tasks. The proposed A-Proactive Grey Wolf Optimization (A-GWO) solves stagnation in GWO by applying a dual search with a Simi-stochastic search. This target is achieved by distributing the population into two groups using a different search technique. The model's performance is evaluated using two benchmarks: the Evolutionary Computation Benchmark (CEC 2005) and seven popular biological datasets. A-GWO demonstrates highly improved efficiency in comparision to the original GWO and Particle Swarm Optimization (PSO). Specifically, it enhances exploration in 66% of CEC functions and achieves high accuracy in 70% of biological datasets.

RViT: Robust Fusion Vision Transformer with Variational Hierarchical Denoising Process for Image Classification

Transformers designed for natural language processing have originally been explored for computer vision in recent research. Various Vision Transformers (ViTs) play an increasingly important role in the field of image tasks such as computer vision, multimodal fusion and multimedia analysis. However, to obtain promising performance, most existing ViTs usually rely on artificially filtered high-quality images, which may suffer from inherent noise risk. Generally, such well-constructed images are not always available in every situation. To this end, we propose a Robust ViT (RViT) to focus on the relevant and robust representation learning for image classification tasks. Specifically, we first develop a novel Denoising VTUnet module, where we conceptualize the nonrobust noise as the uncertainty under the variational conditions. Furthermore, we design a fusion transformer backbone with a tailored fusion attention mechanism to perform image classification based on the extracted robust representations effectively. To demonstrate the superiority of our model, the compared experiments are conducted on several popular datasets. Benefiting from the sequence regularity of the Transformer and captured robust feature, the proposed method exceeds compared Transformer-based models with superior performance in visual tasks.

An example of the application of artificial intelligence models in human resources processes

Creating job postings and selecting suitable candidates among these job postings is a challenging process. This process increases the workload of human resources and causes the process to proceed slowly. It is of great importance for human resources departments to utilize information processing technologies to create job postings effectively and to evaluate the CVs of applicants to these postings. This study introduces and analyzes two different technologies that can help human resources. In the process of preparing job advertisements in the field of IT, in the first stage, the word cloud method is used to decide which keywords should be emphasized in the advertisement texts. In the second stage, the resumes of the applicants are analyzed using three different deep learning models such as CNN (Convolutional Neural Network), GRU (Gated Recurrent Unit), and LSTM (Long Short-Term Memory) for classification purposes. While the performance of these models is evaluated using metrics such as accuracy, MCC, F_1 score, and MSE, the decision-making processes of the models with explainable artificial intelligence are also analyzed. In this context, the GRU model, which achieved an accuracy of 99%, provided the most superior result in this study and the literature. This research shows that deep learning models provide high accuracy rates and efficiency in human resources resume classification and candidate matching processes. It also explains that using the word cloud method, the most appropriate keywords can be identified, and advertisements can be created.

NEURAL NETWORKS AND DEEP LEARNING: ENHANCING AI THROUGH NEURAL NETWORK OPTIMIZATION

Neural networks and deep learning have profoundly impacted artificial intelligence (AI), driving advancements across numerous applications. However, optimizing these networks remains a critical challenge, necessitating sophisticated techniques and methodologies. This article explores the state-of-the-art in neural network optimization, delving into advanced gradient descent variants, regularization methods, learning rate schedulers, batch normalization, and cutting-edge architectures. We discuss their theoretical underpinnings, implementation complexities, and empirical results, providing insights into how these optimization strategies contribute to the development of high-performance AI systems. Case studies in image classification and natural language processing illustrate practical applications and outcomes. The article concludes with an examination of current challenges and future directions in neural network optimization, emphasizing the need for scalable, interpretable, and robust solutions.

Development of a true density-based automated quality grading device for unboiled arecanut kernels

The automated sorting of arecanut kernels is a significant challenge that has not been effectively addressed thus far. Scientific grading techniques are necessary given the paradigm change toward investigating alternative uses for arecanuts in industry and the medical field. This research work emphasizes the relatively unexplored aspect of the post-harvest process; quality grading of kernels based on physical properties. It aimed to develop a novel approach for classifying unboiled (Chali) arecanut kernels cultivated in Goa, India based on their true density, using a combination of mechanical and visual techniques. The study explored the potential of true density as a quality indicator for real-time grading of the kernels. To achieve this, automated grading equipment was devised, utilizing a load cell to measure the kernel's mass and the ellipsoid approximation method to estimate its volume. A machine vision system captured the top and side images of the kernels to measure their volume. Python programs were created to enable image acquisition, processing, object detection, measurement and kernel segregation. Real-time kernel classification was accomplished by establishing serial data communication between the Python code and the Arduino board. The kernel segregation process was facilitated by servomechanism and a stepper motor. The kernels were classified into acceptable and non-acceptable categories based on a threshold value of true density. The research successfully established a method that utilizes the physical attributes of arecanut kernels as parameters for quality grading. However, the study encountered challenges with the density measurements, as the paired t-test results revealed significant differences between the kernel true density measured by the device and the true density estimated using the weighing scale-water displacement method, indicating a percentage error of 13.2%. Addressing these challenges would lead to more accurate density calculations, thereby enhancing the overall effectiveness of the kernel classification process. Furthermore, the technique allowed for the offline estimation of the kernels' porosity, which was found to be 45.3%. In future research, the integration of density and porosity measurement systems could be explored for real-time quality evaluation based on porosity, offering potential opportunities for further enhancement and optimization of the grading process. The technology could be further applied to other types of nuts and agricultural products, thereby overcoming the limitations of color-based sorting using image processing. Key words: Quality grading, True density, Machine vision, Arecanut kernel, Python

Advancing allergy source mapping: A comprehensive multidisciplinary framework integrating machine learning, graph theory and game theory

Allergic reactions can range from mild discomfort to life-threatening situations. To manage the healthcare difficulty, an efficient allergens mapping is required by mapping the allergies to reduce the severity risk reactions. The allergies are mapping with specific food items according to the daily usage. It enables the clear communication, targeted avoidance, and public awareness, which is necessary to scrutinize the prevention process. Since, the allergies are highly individualized and can exhibit significant variability among individuals that are capturing the complexity in the process of overcoming the predictive challenges. The complex relationships and data challenges require advanced approaches like ML and graph theory. For this purpose, we propose a new multidisciplinary framework that integrates the Machine Learning (ML), Graph Theory and Game Theory to predict the allergies associated with relevant foods using a modest dataset. This framework has two newly built graph models such as a conventional approach and a refined approach, to pave the way for better results. Here, the ML techniques are employed to perform the classification process on tabular data that are observed the remarkable improvements and transforming the data into a graph. Moreover, the Darwinian decision-making framework is adopted theoretically in evolutionary game theory to formulate effective formulas for assessing the spread of allergies among allergens and predict the allergies dynamically. The proposed framework has been evaluated by conducting experiments by using a modest dataset by considering the evaluation metrics such as accuracy, macro-precision, macro-recall, and macro F1-score.

Automated Crack Detection in Monolithic Zirconia Crowns Using Acoustic Emission and Deep Learning Techniques.

Monolithic zirconia (MZ) crowns are widely utilized in dental restorations, particularly for substantial tooth structure loss. Inspection, tactile, and radiographic examinations can be time-consuming and error-prone, which may delay diagnosis. Consequently, an objective, automatic, and reliable process is required for identifying dental crown defects. This study aimed to explore the potential of transforming acoustic emission (AE) signals to continuous wavelet transform (CWT), combined with Conventional Neural Network (CNN) to assist in crack detection. A new CNN image segmentation model, based on multi-class semantic segmentation using Inception-ResNet-v2, was developed. Real-time detection of AE signals under loads, which induce cracking, provided significant insights into crack formation in MZ crowns. Pencil lead breaking (PLB) was used to simulate crack propagation. The CWT and CNN models were used to automate the crack classification process. The Inception-ResNet-v2 architecture with transfer learning categorized the cracks in MZ crowns into five groups: labial, palatal, incisal, left, and right. After 2000 epochs, with a learning rate of 0.0001, the model achieved an accuracy of 99.4667%, demonstrating that deep learning significantly improved the localization of cracks in MZ crowns. This development can potentially aid dentists in clinical decision-making by facilitating the early detection and prevention of crack failures.

Rapid CU Partitioning and Joint Intra-Frame Mode Decision Algorithm

H.266/Versatile Video Coding (VVC) introduces new techniques that build upon previous standards, proposing a nested multi-type tree quadtree (QTMT). The introduction of this structure significantly enhances video coding efficiency; additionally, the number of directional modes in H.266 has increased by 32 compared to H.265, accommodating a greater variety of texture patterns. However, the changes in the related structures have also led to a significant increase in encoding complexity. To address the issue of excessive computational complexity, this paper proposes a targeted rapid Coding Units segmenting approach combined with decision-making for an intra-frame modes algorithm. In the first phase of the algorithm, we extract different features for CU blocks of various sizes and input them into the decision tree model’s classifier for classification processing, determining the CU partitioning mode to prematurely terminate the partitioning, thereby reducing the encoding complexity to some extent. In the second phase of the algorithm, we put forward an intra-frame mode decision strategy grounded in gradient descent techniques with a bidirectional search mode. This maximizes the approach to the global optimum, thereby obtaining the optimal intra-frame mode and further reducing the encoding complexity. Experimentation has demonstrated that the algorithm achieves a 54.53% reduction in encoding time. In comparison, the BD-BR (Bitrate-Distortion Rate) only increases by 1.38%, striking an optimal balance between the fidelity of video and the efficacy of the encoding process.

Comparative Analysis for Bag of Features (BoF) Performance

The accuracy of the Bag of Features (BoF) is greatly affected by the discriminatory power of feature extraction techniques. This paper presented a proposed method designed to find the best features technique for constructing a BoF model according to the image classification. It consists of four stages: feature extraction, where detectors and descriptor feature techniques have been exploited to generate different BoF models. Each BoF model is generated depending on what detector and descriptor are used. The BoF models are constructed to represent the images as feature vectors. The classification process is then performed on two image datasets. Finally, the efficiency of BoF models is analyzed and evaluated with respect to the accuracy of their classification performance. Experimental results indicated that the best level of accuracy was provided by the proposed BoF model with the KAZE features method. The results also showed that the BoF model with Speeded Up Robust Features (SURF) was superior to other feature methods in terms of execution time, which was 0.01218 seconds. Moreover, the BoF model generated by the SURF detector combined with the KAZE descriptor achieved a high level of accuracy of 0.99 and kept the time complexity low (0.01948 seconds).

Optimizing agricultural land use: A GIS-based assessment of suitability in the Sana River Basin, Bosnia and Herzegovina

Abstract The research subject is the application of geographic information systems (GIS) in assessing land suitability for agriculture in the Sana River Basin in Bosnia and Herzegovina. The aim of the research is to apply the analytic hierarchy process (AHP) in identifying suitable areas for agricultural production. Within the AHP framework, the following factors were considered: pedology, land use, elevation, slope, aridity index, and distance from rivers. The results of the suitability assessment underwent reclassification (RP) and cluster analysis processes (CAPs). It was found that very unsuitable land (1) covers an area of 0.15% (RP) or 5.83% (CAP), unsuitable land (2) covers 3.44% (RP) or 17.52% (CAP), conditionally suitable land (3) covers 32.11% or 28.47% (CAP), suitable land (4) covers 56.29% or 28.57% (CAP), and very suitable land (5) covers 7.98% (RP) or 19.59% (CAP). At the study area level, a supervised classification process was conducted to identify land use classes: meadows/pastures, water, forest, agricultural, and built-up areas. RP and CAP results were overlaid with supervised classification results to determine the amount of land used for agricultural purposes within each suitability class. It was determined that currently, for agricultural purposes, 0.04 km2 (RP) or 0.88 km2 (CAP) of very unsuitable land (1) is used, 0.41 km2 (RP) or 7.28 km2 (CAP) of unsuitable land (2), 15.75 km2 (RP) or 27.52 km2 (CAP) of conditionally suitable land (3), 185.15 km2 (RP) or 107.06 km2 (CAP) of suitable land (4), and 42.99 km2 (RP) or 101.65 km2 (CAP) of very suitable land (5). The research findings hold substantial importance in elucidating both the potential and constraints of land use practices as a vital natural resource within agriculture. They also have practical importance for relevant institutions in terms of agricultural sector development and making timely land use planning decisions for sustainable development.

Change detection due to ISPAAD Programme using geospatial techniques: A case study of Dinogeng Agricultural Extension Area of Kgatleng District, Botswan

Dinogeng Agricultural Extension Area (DAEA), located in the south eastern part of Botswana has witnessed tremendous land use changes due to the Integrated Support Programme for Arable Agriculture Development (ISPAAD). For over a decade, crop cultivation has been relatively small both in size and population. But today, Dinogeng is one of the fastest growing agricultural areas. Therefore, it is paramount to detect the nature and magnitude of land use changes for planning purpose. Remotely Sensed data from Landsat 5 and Landsat 8 were utilized for the purpose of Land Use Land Cover (LULC) change detection. Geographical Information Systems (GIS) and Remote Sensing (RS) were used to produce LULC maps for 2006 and 2020 for assessing the severity of land degradation. In a 14-year span (2006-2020), LULC of DAEA changed markedly. Cultivated land and bare areas increased by 19.4 and 18.3 % whereas shrub land and forest areas decreased by 36.9 and 0.7 %, respectively. Supervised classification algorithms and stratified random sampling design were adopted for the accuracy assessment. The classification process produced good results with overall accuracies of 93% and 94% for the 2006 and 2020 maps, respectively. The findings could be useful to guide the development of functional land use plan for DAEA.

An algorithm for building contour inference fitting based on multiple contour point classification processes

Extracting buildings from True Digital Ortho Maps often encounters occlusions and misidentifications, making it challenging to obtain complete, regular, and accurate building contours. To address this issue, we developed a building recognition process based on the Segment Anything Model, and proposed a novel regularization algorithm for building contour inference and fitting, which quantifies the confidence levels of contour points to accurately fit building contours from data containing substantial noise, and reformulates the fitting problem as progressive node classification tasks consisting of contour simplification, iterative regularization, and rationality assessment. In experimental evaluations, the proposed contour fitting algorithm achieved 97.99 % Intersection over Union (IoU), 95.39 % consistency with the standard contour edge count, and 88.06 % of cases with Hausdorff distances less than or equal to 15 pixels (30 cm), significantly outperforming comparative methods. Notably, it was the only contour regularization algorithm that improved IoU (1.03 %) compared to the original contours. The experimental results demonstrate that the proposed algorithm effectively suppresses noise and infers incomplete building contours, producing accurate and regular contours comparable to manual delineation. It is particularly suitable for buildings with near-orthogonal structures, exhibiting significant practical applicability and generalization potential.

AUTORECYCLER: Prototype based on artificial vision to automate the material classification process (Plastic, Glass, Cardboard and Metal)

Environmental protection has gained greater importance over time due to the negative impact and irreversible consequences that have occurred worldwide and stem from pollution. One of the great challenges faced in different parts of the world is the inadequate management and classification of solid waste. In order to contribute to tackling this issue, this paper proposes an automated sorting system based on artificial vision which allows recognition and separation of recyclable materials (Plastic, Glass, Cardboard and Metal) through a webcam connected in real time to the Nvidia® Jetson Nano™ 2 GB programming board, which has a convolutional neural network (CNN) trained for the proper classification of waste. The system had a 95 % accuracy in separating plastic, 96 % in glass and metal, and 94 % in cardboard. With this in mind, we conclude it contributes to the recycling effort, which has an impact on the reduction of environmental pollution worldwide.

Learning and using multiple languages: Experiences of adults with ADHD

Attention-Deficit/Hyperactivity Disorder (ADHD) is a neurodevelopmental disorder that impacts many areas of a person's life. So far little is known on how ADHD affects the learning and usage of additional languages. The present study investigates what potential challenges and benefits adults with ADHD experience in second language (L2) learning, and what strategies they have developed to offset ADHD-related differences in cognition and behaviour. As part of a larger questionnaire, 59 participants with ADHD provided written responses, answering questions on how ADHD has influenced their learning and use of multiple languages. After a multiple-step classification process, three overarching topics were identified: (1) the effects of cognition on language learning and use; (2) how different language components and language modalities are affected by ADHD; and (3) what language learning strategies adults with ADHD apply. In our analysis, we present authentic quotes from individuals with ADHD and place them into the larger context of research on neurodivergence and L2 acquisition. Our results highlight the great heterogeneity in learning experiences and pathways in adults with ADHD. Interestingly, many respondents embrace their neurodiversity and develop individual strategies to enhance L2 proficiency. Our research aims to contribute to a better understanding of multilingualism within neurodiverse populations.