Real-world Applications Research Articles

Empirical Likelihood Ratio Based K-Nearest Neighbours Regression

Regression models play a pivotal role in real-life applications by enabling the analysis and prediction of continuous outcomes. Among these, the k-Nearest Neighbours (KNN) model stands out as a significant advancement in machine learning. KNN's ability to make predictions based on the proximity of data points has found wide-ranging applications in various fields. However, the traditional KNN regression model has its limitations, including sensitivity to noise and an uneven distribution of neighbours. In response, this paper introduces a novel approach: an empirical likelihood ratio (ELR) based regression algorithm. The ELR technique offers distinct advantages over distance-based nearest neighbour computations, particularly in handling skewed data distributions and minimizing the impact of outliers. The proposed ELRbased KNN regression model is rigorously assessed through both simulation studies and real-life scenarios. The results explicitly demonstrate the enhanced performance of the ELR-based approach over the conventional KNN model. This research contributes to a deeper understanding of regression techniques and underscores the practical significance of leveraging empirical likelihood ratios in refining predictive models for real-world applications.. KEYWORDS :Regression model, k-Nearest neighbours, Empirical likelihood ratio, Distance measures, Data distributions.

Synthesis and characterization of a copper-based metal–organic framework and its application in microextraction and determination of chlorpyrifos by ion mobility spectrometry

This study describes an innovative method for the microextraction and determination of chlorpyrifos, utilizing a newly synthesized copper-based metal–organic framework (MOF) with 3,3ʹ-thiodipropionic acid as the ligand. This approach offers a sensitive and reliable solution for the ultra-trace determination of chlorpyrifos in various matrices, highlighting its potential for broader environmental and food safety applications. To minimize environmental impact, a hydrothermal method was chosen for MOF synthesis. The MOF was fully characterized using FT-IR, X-ray diffraction, FESEM-EDX, BET, zeta potential analysis, and XPS; and integrated into a dispersive solid-phase microextraction (DSPME) technique, paired with ion mobility spectrometry (IMS) for chlorpyrifos detection. The DSPME procedure has been optimized for extraction efficiency, minimal solvent use, and low absorbent consumption. A low detection limit of 0.65 ng mL−1 was observed and validated along with a linear detection range of 1.0–400.0 ng mL−1 as the result of this optimization. Experiments with water, pear fruit, and soil samples demonstrated the method’s practicality and superior performance compared to existing techniques, showcasing its potential for real-world applications.

Green electrochemical sensor for selective determination of anticancer drug Ribociclib based on banana peel activated carbon/CuCoFe-MOF/CoFe2O4 modified carbon paste electrode in pharmaceutical and biological samples

The development of efficient and eco-friendly sensing platforms for the sensitive detection of anticancer drugs like Ribociclib (RIB) is essential for advancing therapeutic monitoring and ensuring patient safety. In this study, we introduce a novel electrochemical sensor that utilizes a unique CuCoFe-MOF/CoFe2O4/banana peel activated carbon (BPAC) composite-modified carbon paste electrode (CPE) for the sensitive quantification of RIB. This composite material, which has not been previously reported in the literature, offers enhanced electrocatalytic activity, attributed to its increased surface area and superior conductivity. Various techniques, including FT-IR, XRD, SEM-EDX, and BET analysis, were used to confirm the structural and morphological properties of the synthesized composite. Differential pulse voltammetry (DPV) results showed a broad linear range of 0.2–9.7 μM and a remarkably low limit of detection (LOD) of 0.025 μM, demonstrating its high sensitivity. The sensor’s performance was further validated in pharmaceutical formulations and biological samples, achieving recoveries between 98.6 % and 101.8 % with an RSD below 2.0 %. Additionally, the method was evaluated using the Green Analytical Procedure Index (GAPI), Analytical Greenness (AGREE), and Blue Applicability Grade Index (BAGI) tools, confirming that it is both green and highly applicable. This work presents a significant advancement in the field of electrochemical sensing for RIB detection, offering a novel, sustainable, and practical solution for real-world applications.

A novel hydrodynamic semiconductor model under Hall current effect and laser pulsed excitation

This paper presents a novel investigation into the effects of a strong external magnetic field on a hydro-poroelastic semiconductor model within the framework of photo-thermoelasticity theory in two dimensions. This study focuses on generating a Hall current and its impact on the coupled behavior of thermal, mechanical, and electronic fields in a semiconductor medium saturated with fluid. Using the normal mode analysis, we derive and analyze the wave propagation characteristics of key physical fields, including the non-dimensional temperature, displacement, mechanical stresses, carrier density, and excess pore water pressure, in response to the imposed magnetic field. Boundary conditions relevant to real-world applications are incorporated to assess the interactions between these fields. The results provide insight into the dynamic coupling of electromagnetic, thermal, and mechanical phenomena in porous semiconductor materials and offer potential applications in the design of magneto-sensitive semiconductor devices as well as in geophysical and biomedical engineering fields where such multi-field interactions are critical. The results obtained are plotted with some comparisons.

A Survey on Efficient Vision Transformers: Algorithms, Techniques, and Performance Benchmarking.

Vision Transformer (ViT) architectures are becoming increasingly popular and widely employed to tackle computer vision applications. Their main feature is the capacity to extract global information through the self-attention mechanism, outperforming earlier convolutional neural networks. However, ViT deployment and performance have grown steadily with their size, number of trainable parameters, and operations. Furthermore, self-attention's computational and memory cost quadratically increases with the image resolution. Generally speaking, it is challenging to employ these architectures in real-world applications due to many hardware and environmental restrictions, such as processing and computational capabilities. Therefore, this survey investigates the most efficient methodologies to ensure sub-optimal estimation performances. More in detail, four efficient categories will be analyzed: compact architecture, pruning, knowledge distillation, and quantization strategies. Moreover, a new metric called Efficient Error Rate has been introduced in order to normalize and compare models' features that affect hardware devices at inference time, such as the number of parameters, bits, FLOPs, and model size. Summarizing, this paper first mathematically defines the strategies used to make Vision Transformer efficient, describes and discusses state-of-the-art methodologies, and analyzes their performances over different application scenarios. Toward the end of this paper, we also discuss open challenges and promising research directions.

A resilient biofuel supply chain design based on Paulownia and Jatropha under uncertainty considering the water-energy nexus

The growth in global population, pollution, and energy demand, in addition to dwindling fossil fuel reserves, are driving scientists and industries toward adopting biomass-based products over non-renewable energy sources. Moreover, the decline of water resources and the importance of conserving water have induced governments to start initiatives to control the consumption of water, especially in the agricultural and industrial sectors. Therefore, the water-energy nexus is a significant issue that has drawn the interest of researchers in recent years. A bi-objective two-stage stochastic multi-product optimization model is presented for designing a resilient second-generation biofuel supply chain under uncertainty, considering the water-energy nexus. In this supply chain, biofuel is yielded from inedible plants Paulownia and Jatropha. Various resilience strategies are adopted to cope with potential disruptions to the supply chain, including multi-sourcing, capacity expansion, imports, multiple transportation modes, and lateral transshipment. The objectives of the model are twofold: maximizing profit and minimizing water consumption. This is achieved by making some key decisions including: facility location, managing the flows of raw materials and finished products among networks, determining production volumes, setting inventory levels, allocating land for biomass crop cultivation, planning production and storage capacities, scaling capacity expansions, the number of vehicles to rent or purchase, and optimizing the amounts of raw materials to import and biomass to export. An actual case study is provided to demonstrate the real-world applicability of the approach, while multiple sensitivity analyses are performed on the problem's main parameters. The study concludes by providing valuable managerial insights that focus on the cost savings and profit increases achievable through resilience strategies. The analysis considers different scenarios and explores effective methods for handling disruptions, emphasizing water resource conservation achieved by integrating the water-energy nexus within the biofuel supply chain.

Automated Non-Supervised Eye Disorder Screening System Using Virtual Reality.

Neglecting eye disorders can lead to visual impairment or even vision loss, making early diagnosis and medical treatment crucial. This paper presents the development of an automated non-supervised eye disorder screening system that utilises Virtual Reality (VR) to provide a preliminary screening for common eye disorders. The system integrates advanced pupil-tracking techniques, image-processing methods, and eye disorder screening algorithms into one package for comprehensive eye disorder monitoring, including a strabismus test, a pupil test, and a contrast sensitivity test. Patients can wear a VR headset and use a joystick to interact with the VR scenario. The system is validated through experiments conducted on both healthy subjects and patients in hospitals, demonstrating an agreement rate exceeding 90% when compared to diagnoses by doctors. The test results highlight its potential for real-world applications, ultimately improving the accessibility of early screening for eye disorders and remote healthcare.

Automated Machine Learning: A Survey of Tools and Techniques

Automated Machine Learning (AutoML) has revolutionized the field of machine learning by automating complex and time-intensive tasks such as data preprocessing, model selection, and hyperparameter tuning. This study explores the capabilities, limitations, and practical applications of six widely used AutoML tools: Auto-sklearn, TPOT, H2O.ai, Google Cloud AutoML, Microsoft Azure AutoML, and Amazon SageMaker Autopilot. By evaluating these tools across diverse datasets—spanning tabular data, time series, image classification, and text sentiment analysis—the research highlights their predictive performance, computational efficiency, scalability, and explainability. Proprietary tools demonstrated superior scalability and efficiency through cloud integration, while open-source platforms provided more interpretability and flexibility. However, challenges such as lack of transparency in advanced neural architecture search mechanisms and ethical considerations, including bias mitigation, remain prevalent. This study concludes that while AutoML tools significantly lower the barrier to entry for machine learning, ongoing advancements are required to balance performance, usability, and ethical standards, making AutoML an integral solution for real-world applications.

Pectin/zinc alginate films containing anthocyanins from dragon fruit peel as intelligent pH indicators for shrimp freshness monitor

The novel incorporation of dragon fruit peel extract (DE), rich in anthocyanins, Zn2+ (from Zinc Alginate) and pectin was applied to create active and intelligent food packaging composite films. These films were characterized for their microstructure and properties. Various levels of anthocyanin extracts (1 %, 3 %, and 5 %) were evaluated for their impact on the films' physical and functional properties, incorporating microstructure, mechanical strength, barrier properties, pH sensitivity, and bacteriostatic effectiveness. The films exhibited a significant antibacterial rate of up to 99.99 % against common foodborne pathogens, enhanced flame retardancy with an enhancement of 32.7 %, and a broad pH sensitivity range, indicating their adaptability to various conditions. The results demonstrated that the prepared indicator film achieved a 50 % reduction in water vapor permeability. Additionally, the mechanical properties were enhanced, with only a slight decrease of 12.2 % in tensile strength and 14.0 % in elongation at break. In tests monitoring shrimp freshness, pectin/ZA/DE films showed notable color changes correlating with shrimp quality. These specific values highlight the pectin/ZA/DE films' potential for real-world applications, suggesting that they have potential applications as smart packaging materials in the food industry.

New order-dependent conditions to control a class of nonlinear real-order systems

In many application points of interest, controlling the responses of real-world applications that vary with time seems quite challenging and puzzling. A linear time-varying state feedback controller is widely known and can be useful to estimate bounds to the state control matrix in the design of many control systems. The issue of initialization of real-order control system enhancement remains a challenging issue in the systems analysis subject to random initial-time placed on a real number line. We address a new design of a class of real-order control systems that consists of separated linear and nonlinear terms affected by input functions to be controlled with an implemented time-varying linear state feedback controller. We utilize the fractional comparison method and under Lipschitz nonlinearity with a constant bounding matrix of time-varying coefficients of control systems to address new order-dependent conditions that provide local and global stabilization to controlled systems. Applications of results that include practical real-order single-machine-infinite-bus power systems have been illustrated to control the responses by the utility of theoretical conditions examined along with validation of numerical simulations. It is shown that the proposed controller is practically convenient and demonstrates the efficiency of measuring the performances of control systems.

Transformer-Based Visual Segmentation: A Survey.

Visual segmentation seeks to partition images, video frames, or point clouds into multiple segments or groups. This technique has numerous real-world applications, such as autonomous driving, image editing, robot sensing, and medical analysis. Over the past decade, deep learning-based methods have made remarkable strides in this area. Recently, transformers, a type of neural network based on self-attention originally designed for natural language processing, have considerably surpassed previous convolutional or recurrent approaches in various vision processing tasks. Specifically, vision transformers offer robust, unified, and even simpler solutions for various segmentation tasks. This survey provides a thorough overview of transformer-based visual segmentation, summarizing recent advancements. We first review the background, encompassing problem definitions, datasets, and prior convolutional methods. Next, we summarize a meta-architecture that unifies all recent transformer-based approaches. Based on this meta-architecture, we examine various method designs, including modifications to the meta-architecture and associated applications. We also present several specific subfields, including 3D point cloud segmentation, foundation model tuning, domain-aware segmentation, efficient segmentation, and medical segmentation. Additionally, we compile and re-evaluate the reviewed methods on several well-established datasets. Finally, we identify open challenges in this field and propose directions for future research.

Enhancing power quality in wireless DC power supplies through active power filtering: A computer simulation approach

This paper presents a computer simulation model for a high-power factor wireless DC power supply system, integrating an active power filter (APF) at the rectifier stage on the transmitter side using a rectifier boost technique. The APF, employing a MOSFET switch regulated by active pulse width modulation (APWM) within a current control loop, addresses pulsating and distorted AC supply currents caused by non-linear loads. A robust closed-loop control mechanism, including a subtractor circuit, proportional-integral (PI) controller, and comparator, ensures the generation of a continuous sinusoidal waveform synchronized with the supply voltage. The model utilizes a high-frequency inverter to convert DC to AC, which is then wirelessly transmitted via wireless power transfer (WPT) technology and converted back to DC by a high-frequency rectifier. MATLAB/Simulink simulation results show a significant reduction in total harmonic distortion (THD) of the AC supply current, complying with IEEE 519 standards. Selected results are presented to verify the proposed method's effectiveness in reducing harmonic distortions and enhancing power quality. This study highlights the advantages of WPT in scenarios where traditional wired connections are impractical and underscores the potential of this system for real-world applications, particularly in developing high-power factor wireless DC power supply systems.

Learning to Cut via Hierarchical Sequence/Set Model for Efficient Mixed-Integer Programming.

Cutting planes (cuts) play an important role in solving mixed-integer linear programs (MILPs), which formulate many important real-world applications. Cut selection heavily depends on (P1) which cuts to prefer and (P2) how many cuts to select. Although modern MILP solvers tackle (P1)-(P2) by human-designed heuristics, machine learning carries the potential to learn more effective heuristics. However, many existing learning-based methods learn which cuts to prefer, neglecting the importance of learning how many cuts to select. Moreover, we observe that (P3) what order of selected cuts to prefer significantly impacts the efficiency of MILP solvers as well. To address these challenges, we propose a novel hierarchical sequence/set model (HEM) to learn cut selection policies. Specifically, HEM is a bi-level model: (1) a higher-level module that learns how many cuts to select, (2) and a lower-level module-that formulates the cut selection as a sequence/set to sequence learning problem-to learn policies selecting an ordered subset with the cardinality determined by the higher-level module. To the best of our knowledge, HEM is the first data-driven methodology that well tackles (P1)-(P3) simultaneously. Experiments demonstrate that HEM significantly improves the efficiency of solving MILPs on eleven challenging MILP benchmarks, including two Huawei's real problems.

Utilizing minimum spanning trees for effective mobile sink routing in wireless sensor networks

With many practical applications, wireless sensor networks (WSNs) represent an important field of study. Real-world applications of WSNs include smart home automation, healthcare, agriculture, industrial automation, and environmental monitoring. WSNs present countless chances for creative solutions across various industries as they develop and become more sophisticated. But because they are unattended, we must devise ways to make them work better without using the sensor nodes’ most important resource—battery power. A unique sink mobility model from a deployed WSN is proposed in this paper, based on constructing a minimal Spanning tree. The proposed approach derives a controlled movement model for the mobile sink based on minimal spanning tree (MST) features. Consequently, fixed nodes will be scheduled and visited to save routing overhead and improve network efficiency. Using the properties of the minimal spanning tree, the moving sink node can visit immobile sensor nodes, which is the most effective approach to gather data and send it to the base station. The effectiveness of WSNs was examined when implementing this mobility model, and we used the NS-2 simulator to run simulations to assess how efficiently the suggested strategy performed. Our findings demonstrate that WSN performance can be significantly enhanced by implementing the proposed method.

3D path planning for UAV based on A hybrid algorithm of marine predators algorithm with quasi-oppositional learning and differential evolution

3D path planning is a critical requirement for the autonomy of unmanned aerial vehicle (UAV) navigation systems. In this paper, we propose a novel approach, the differential evolution combined marine predators algorithm (DECMPA), specifically tailored to address the UAV 3D path planning problem in complex scenarios. To handle stringent constraints, DECMPA adopts a multi-step approach. Initially, it utilizes quasi-oppositional learning to generate a more uniformly distributed initial population. Throughout the iterative process, it probabilistically generates quasi-oppositional populations and merges them with existing populations, selecting the superior individuals for the next generation to mitigate local optima. Additionally, DECMPA integrates the variance, crossover, and selection processes of differential evolution into the marine predators algorithm iterations. The greedy selection of test and target vectors further accelerates population convergence. Numerical optimization experiments are conducted using the 10 and 20-dimensional CEC 2021 test suite. Compared to other algorithms, DECMPA exhibits superior solution accuracy and convergence speed, resulting in substantial performance enhancement, thus establishing itself as an efficient algorithm. Furthermore, the adoption of spherical vector coordinates in solution encoding ensures higher quality and facilitates the production of feasible solutions. To verify the effectiveness and practicality of the proposed algorithm, comprehensive path planning simulation experiments are conducted. DECMPA is compared against nine other state-of-the-art heuristic algorithms across six 3D scenarios of varying complexity. The experimental results demonstrate DECMPA’s superiority in terms of optimal cost acquisition, solution quality, and stability. In conclusion, DECMPA presents a promising solution for addressing the challenges of UAV 3D path planning in complex environments. Its innovative approach and superior performance underscore its potential for real-world application in autonomous UAV navigation systems.

Starch-based encapsulation to enhance probiotic viability in simulated digestion conditions

This research aims to meet the demand for efficient delivery systems in the food, nutraceutical, and pharmaceutical industries. The study involved the synthesis of starch-based nanoparticles for potential application in the encapsulation of Lactobacillus rhamnosus. Various techniques such as zeta sizer, Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), and scanning electron microscopy (SEM) were used to characterize the encapsulated probiotics in microbeads. The results showed 85.00 % encapsulation efficiency of beads. Microscopic analysis revealed that the probiotics accumulated within the wall material and formed small, smooth polygonal granules on the capsule surface. XRD analysis confirmed the presence of amorphous humps and some crystallinity of nanoparticles in the capsules. Moreover, encapsulation significantly improved probiotic viability under simulated gastrointestinal conditions. This study highlights the potential of starch-based nanoparticles to enhance the stability and viability of probiotics, demonstrating their potential applications across various industrial sectors. Further research should focus on investigating the long-term stability and functional efficacy of encapsulated probiotics in microbeads for real-world applications.

Propagation dynamics of vacillating individual with behavioral variation on double-layer social network

This paper investigates behavior propagation in a two-layer network, emphasizing the roles of behavioral vacillation and dependency psychology. By integrating a threshold function model and percolation theory, the study explores how varying degrees of vacillation and network viscosity impact behavior spread. The results show that higher behavioral vacillation accelerates propagation, reduces uncertainty, and lowers the critical threshold, highlighting the significant influence of psychological factors on diffusion. These findings provide valuable insights for real-world applications in marketing strategies, social influence, and public opinion dynamics. Future work should incorporate dynamic network elements to offer a more comprehensive understanding of behavior propagation.

Beat the heat: wearable-based study of perceived heat stress and physiological strain in swiss track workers in a controlled climate chamber

Increasing temperatures pose new challenges for track workers (TWs), who endure prolonged exposure to extreme heat and humidity. New methods are critically needed to assess their performance and heat tolerance, aiming to mitigate workplace accidents and long-term health consequences. This study aimed to investigate the physiological effects of heat exposure on TWs, using wearable sensors to monitor key physiological parameters under controlled environmental conditions. Nineteen TWs participated in the study, which included two experimental sessions simulating different thermal environments: a typical Swiss summer night and a hot summer day. Participants' core body temperature, heart rate (HR), and skin temperature were monitored using wearable sensors, and physiological indexes were computed. In addition, perceptual strain index (PeSI) and psychomotor vigilance task (PVT) response times were recorded. Statistically significant increases in physiological parameters were observed under hotter conditions. The study identified statistically significant correlations between the PeSI and the physiological strain index and between PeSI and HR. Perceptual scores were consistently higher than the values derived from physiological measurements, suggesting a greater subjective experience of heat strain. The PVT response times were higher on the hotter day, reflecting increased cognitive strain due to heat exposure. The study highlights the critical impact of heat stress on TWs, with statistically significant increases in physiological and cognitive strain under higher temperatures. Future research should focus on real-world applications of heat strain monitoring.

UPPER-RIM NITRO AND LOWER-RIM METHOXY FUNCTIONALIZED CALIX[4]ARENE BASED COLORIMETRIC SENSOR FOR DETECTION OF 2,4,6-TRINITROPHENOL

The development of effective sensors capable of detecting 2,4,6-trinitrophenol (TNP), or picric acid, at extremely low levels is a very interesting field of study for prevention of environmental contamination and terrorist threats. Therefore, in this study, a colorimetric sensor (MNC4) based on calix[4]arene containing a chromophore group was prepared for the detection of TNP. The photophysical interaction of MNC4 with target nitroaromatic compounds (NACs) was evaluated using UV-Vis. spectroscopy. The initial MNC4 solution was colorless, but a distinct color change was observed in the case of TNP, indicating a stronger binding affinity with MNC4. Spectral investigations also confirmed the relatively strong complexation between MNC4 and TNP. The limit of detection (LOD) value for TNP was found to be 520 nM, indicating that MNC4 can detect TNP down to the nanomolar level, which is significantly lower than the permissible level (2.2 µM). The stoichiometric ratio of the complex formed between MNC4 and TNP was determined to be 1:2. The developed sensor enables the colorimetric detection of TNP with the naked eye without the need for any instrumentation, highlighting its potential for real-world applications.

Enhanced scene text recognition using deep learning based hybrid attention recognition network

<span lang="EN-US">The technique of automatically recognizing and transforming text that is present in pictures or scenes into machine-readable text is known as scene text recognition. It facilitates applications like content extraction, translation, and text analysis in real-world visual data by enabling computers to comprehend and extract textual information from images, videos, or documents. Scene text recognition is essential for many applications, such as language translation and content extraction from photographs. The hybrid attention recognition network (HARN), unique technology presented in this research, is intended to greatly improve efficiency and accuracy of text recognition in complicated scene situations. HARN makes use of cutting-edge elements including alignment-free sequence-to-sequence (AFS) module, creative attention mechanisms, and hybrid architecture that blends attention models with convolutional neural networks (CNNs). Thanks to its novel attention processes, HARN is capable of comprehending wide range of scene text components by capturing both local and global context information. Through faster network convergence, shorter training times, and better utilization of computing resources, the suggested technique raises bar for state-of-the-art. HARN’s versatility makes it a good choice for range of scene text recognition applications, including multilingual text analysis and data extraction. Extensive tests are conducted to assess the effectiveness of HARN approach and demonstrate it is ability to greatly influence real-world applications where accurate and efficient text recognition is essential.</span>