Search Results (12)

We present a novel deep generative semi-supervised framework for credit card fraud detection, formulated as a time series classification task. As financial transaction data streams grow in scale and complexity, traditional methods often require large labeled datasets and struggle with time series of irregular sampling frequencies and varying sequence lengths. To address these challenges, we extend conditional Generative Adversarial Networks (GANs) for targeted data augmentation, integrate Bayesian inference to obtain predictive distributions and quantify uncertainty, and leverage log-signatures for robust feature encoding of transaction histories. We propose a composite Wasserstein distance-based loss to align generated and real unlabeled samples while simultaneously maximizing classification accuracy on labeled data. Our approach is evaluated on the BankSim dataset, a widely used simulator for credit card transaction data, under varying proportions of labeled samples, demonstrating consistent improvements over benchmarks in both global statistical and domain-specific metrics. These findings highlight the effectiveness of GAN-driven semi-supervised learning with log-signatures for irregularly sampled time series and emphasize the importance of uncertainty-aware predictions. Full article

Detecting surface cracks on sugarcane is a critical step in ensuring product quality control, with detection precision directly impacting raw material screening efficiency and economic benefits in the sugar industry. Traditional methods face three core challenges: (1) complex background interference complicates texture feature extraction; (2) variable crack scales limit models’ cross-scale feature generalization capabilities; and (3) high computational complexity hinders deployment on edge devices. To address these issues, this study proposes a lightweight sugarcane surface crack detection model, SCS-YOLO (Surface Cracks on Sugarcane-YOLO), based on the YOLOv10 architecture. This model incorporates three key technical innovations. First, the designed RFAC2f module (Receptive-Field Attentive CSP Bottleneck with Dual Convolution) significantly enhances feature representation capabilities in complex backgrounds through dynamic receptive field modeling and multi-branch feature processing/fusion mechanisms. Second, the proposed DSA module (Dynamic SimAM Attention) achieves adaptive spatial optimization of cross-layer crack features by integrating dynamic weight allocation strategies with parameter-free spatial attention mechanisms. Finally, the DyHead detection head employs a dynamic feature optimization mechanism to reduce parameter count and computational complexity. Experiments demonstrate that on the Sugarcane Crack Dataset v3.1, compared to the baseline model YOLOv10, our model achieves mAP50:95 to 71.8% (up 2.1%). Simultaneously, it achieves significant reductions in parameter count (down 19.67%) and computational load (down 11.76%), while boosting FPS to 122 to meet real-time detection requirements. Considering the multiple dimensions of precision indicators, complexity indicators, and FPS comprehensively, the SCS—YOLO detection framework proposed in this study provides a feasible technical reference for the intelligent detection of sugarcane quality in the raw materials of the sugar industry. Full article

Workflow scheduling in cloud computing is attracting increasing attention. Cloud computing can assign tasks to available virtual machine resources in cloud data centers according to scheduling strategies, providing a powerful computing platform for the execution of workflow tasks. However, developing effective workflow scheduling algorithms to find optimal or near-optimal task-to-VM allocation solutions that meet users’ specific QoS requirements still remains an open area of research. In this paper, we propose a hybrid QoS-aware workflow scheduling algorithm named HLWOA to address the problem of simultaneously minimizing the completion time and execution cost of workflow scheduling in cloud computing. First, the workflow scheduling problem in cloud computing is modeled as a multi-objective optimization problem. Then, based on the heterogeneous earliest finish time (HEFT) heuristic optimization algorithm, tasks are reverse topologically sorted and assigned to virtual machines with the earliest finish time to construct an initial workflow task scheduling sequence. Furthermore, an improved Whale Optimization Algorithm (WOA) based on Lévy flight is proposed. The output solution of HEFT is used as one of the initial population solutions in WOA to accelerate the convergence speed of the algorithm. Subsequently, a Lévy flight search strategy is introduced in the iterative optimization phase to avoid the algorithm falling into local optimal solutions. The proposed HLWOA is evaluated on the WorkflowSim platform using real-world scientific workflows (Cybershake and Montage) with different task scales (100 and 1000). Experimental results demonstrate that HLWOA outperforms HEFT, HEPGA, and standard WOA in both makespan and cost, with normalized fitness values consistently ranking first. Full article

Priority in task scheduling and resource allocation for cloud computing has attracted significant attention from the research community. However, traditional scheduling algorithms often lack the ability to differentiate between tasks with varying levels of importance. This limitation presents a challenge when cloud servers must handle diverse tasks with distinct priority classes and strict quality of service requirements. To address these challenges in cloud computing environments, particularly within the infrastructure of service models, we propose a novel, self-adaptive, multiclass priority algorithm with VM clustering for resource allocation. This algorithm implements a four-tiered prioritization system to optimize key objectives, including makespan and energy consumption, while simultaneously optimizing resource utilization, degree of imbalance, and waiting time. Additionally, we propose a resource prioritization and load-balancing model based on the clustering technique. The proposed work was validated through multiple simulations using the CloudSim simulator, comparing its performance against well-known task scheduling algorithms. The simulation results and analysis demonstrate that the proposed algorithm effectively optimizes makespan and energy consumption. Specifically, our work achieved percentage improvements ranging from +0.97% to +26.80% in makespan and +3.68% to +49.49% in energy consumption while also improving other performance metrics, including throughput, resource utilization, and load balancing. This novel model demonstrably enhances task scheduling and resource allocation efficiency, particularly in complex scenarios with tight deadlines and multiclass priorities. Full article

The emergence of the Internet of Things (IoT) has attracted significant attention in industrial environments. These applications necessitate meeting stringent latency and reliability standards. To address this, the IEEE 802.15.4e standard introduces a novel Medium Access Control (MAC) protocol called Time Slotted Channel Hopping (TSCH). Designing a centralized scheduling system that simultaneously achieves the required Quality of Service (QoS) is challenging due to the multi-objective optimization nature of the problem. This paper introduces a novel optimization algorithm, QoS-aware Multi-objective enhanced Differential Evolution optimization (QMDE), designed to handle the QoS metrics, such as delay and packet loss, across multiple services in heterogeneous networks while also achieving the anticipated service throughput. Through co-simulation between TSCH-SIM and Matlab, R2023a we conducted multiple simulations across diverse sensor network topologies and industrial QoS scenarios. The evaluation results illustrate that an optimal schedule generated by QMDE can effectively fulfill the QoS requirements of closed-loop supervisory control and condition monitoring industrial services in sensor networks from 16 to 100 nodes. Through extensive simulations and comparative evaluations against the Traffic-Aware Scheduling Algorithm (TASA), this study reveals the superior performance of QMDE, achieving significant enhancements in both Packet Delivery Ratio (PDR) and delay metrics. Full article

Drones are used in complex scenes in different scenarios. Efficient and effective algorithms are required for drones to track targets of interest and protect allied targets in a versus game. This study used physical models of quadcopters and scene engines to investigate the resulting performance of attacker drones and defensive drones based on deep reinforcement learning. The deep reinforcement learning network soft actor-critic was applied in association with the proposed reward and penalty functions according to the design scenario. AirSim UAV physical modeling and mission scenarios based on Unreal Engine were used to simultaneously train attacking and defending gaming skills for both drones, such that the required combat strategies and flight skills could be improved through a series of competition episodes. After 500 episodes of practice experience, both drones could accelerate, detour, and evade to achieve reasonably good performance with a roughly tie situation. Validation scenarios also demonstrated that the attacker–defender winning ratio also improved from 1:2 to 1.2:1, which is reasonable for drones with equal flight capabilities. Although this showed that the attacker may have an advantage in inexperienced scenarios, it revealed that the strategies generated by deep reinforcement learning networks are robust and feasible. Full article

Unmanned Aerial Vehicle (UAV) aerial sensors are an important means of collecting ground image data. Through the road segmentation and vehicle detection of drivable areas in UAV aerial images, they can be applied to monitoring roads, traffic flow detection, traffic management, etc. As well, they can be integrated with intelligent transportation systems to support the related work of transportation departments. Existing algorithms only realize a single task, while intelligent transportation requires the simultaneous processing of multiple tasks, which cannot meet complex practical needs. However, UAV aerial images have the characteristics of variable road scenes, a large number of small targets, and dense vehicles, which make it difficult to complete the tasks. In response to these issues, we propose to implement road segmentation and on-road vehicle detection tasks in the same framework for UAV aerial images, and we conduct experiments on a self-constructed dataset based on the DroneVehicle dataset. For road segmentation, we propose a new algorithm C-DeepLabV3+. The new algorithm introduces the coordinate attention (CA) module, which can obtain more accurate segmentation target location information and make the segmentation target edges more continuous. Also, the improved algorithm introduces the cascade feature fusion module to prevent the loss of detail information in road segmentation and to obtain better segmentation performance. For vehicle detection, we propose an improved algorithm S-YOLOv5 by adding a parameter-free lightweight attention module SimAM. Finally, the proposed road segmentation–vehicle detection framework is utilized to unite the C-DeepLabV3+ and S-YOLOv5 algorithms for the implementation of the serial tasks. The experimental results show that on the constructed ViDroneVehicle dataset, the C-DeepLabV3+ algorithm has an mPA value of 98.75% and an mIoU value of 97.53%, which can better segment the road area and solve the problem of occlusion. The mAP value of the S-YOLOv5 algorithm has an mAP value of 97.40%, which is more than YOLOv5’s 96.95%, which effectively reduces the vehicle omission and false detection rates. By comparison, the results of both algorithms are superior to multiple state-of-the-art methods. The overall framework proposed in this paper has superior performance and is capable of realizing high-quality and high-precision road segmentation and vehicle detection from UAV aerial images. Full article

The 14 geometrical tolerances defined in ISO 1101 are supplemented by the alphanumerical symbols defined in this standard and some other standards. The symbols CZ (combined zone), SZ (separate zones) and SIM (simultaneous requirement), which are crucial for the development of robust measuring programs for coordinate measuring machines, were introduced at different times or in different standards. It is shown that the symbol definitions are not always complete. Sometimes there are no univocal rules for their use, which, in some cases, leads to ambiguity in the specifications given by a designer. It is also pointed out that certain functional requirements can be controlled by different symbols, and it is not always clear if the indications are equivalent. This makes it difficult to understand and interpret a drawing by a metrologist and, thus, may lead to uncertainty in the assessment of product conformity regarding specifications. The identified ambiguities and problems in the specification of functional requirements are shown in several figures. Corrections and additions to current standards are proposed. Full article

Cluster microgrids are a group of interoperable smart microgrids, connected in a local network to exchange their energy resources and collectively meet their load. A microgrid can import/export energy to the neighboring microgrid in the network based on energy deficit/availability. However, in executing such an operation, a well-established communication network is essential. This network must provide a reliable communication path between microgrids. In addition, the network must provide an optimal path between any two microgrids in the network to optimize immediate energy generation, import requirements, and export possibilities. To meet these requirements, different conventional research approaches have been used to provide reliable communication, such as backup/alternative/Hot Standby Router Protocol (HSRP)-based redundant path concepts, in addition to traditional/renowned Dijkstra algorithms, in order to find the shortest path between microgrids. The HSRP-based mechanism provides an additional path between microgrids, but may not completely solve the reliability issue, especially during multiple link failures and simultaneous failures of the actual path and redundant path. Similarly, Dijkstra algorithms discussed in the literature do not work for finding the shortest path during link failures. Thus, to enhance reliability, this paper proposes a refined network topology that provides more communication paths between microgrids, while retaining the same number of total links needed, as in conventional HSRP-based networks. In addition, this paper proposes an enhanced Dijkstra algorithm to find the optimum path during link failures. Simulations are executed using NetSim^TM by implementing test cases such as single-link and multiple-link failures. The results prove that the proposed topology and method are superior to conventional approaches. Full article

A multiple-actuator fault isolation approach for overactuated electric vehicles (EVs) is designed with a minimal ${\ell }_1$-norm solution. As the numbers of driving motors and steering actuators increase beyond the number of controlled variables, an EV becomes an overactuated system, which exhibits actuator redundancy and enables the possibility of fault-tolerant control (FTC). On the other hand, an increase in the number of actuators also increases the possibility of simultaneously occurring multiple faults. To ensure EV reliability while driving, exact and fast fault isolation is required; however, the existing fault isolation methods demand high computational power or complicated procedures because the overactuated systems have many actuators, and the number of simultaneous fault occurrences is increased. The method proposed in this paper exploits the concept of sparsity. The underdetermined linear system is defined from the parity equation, and fault isolation is achieved by obtaining the sparsest nonzero component of the residuals from the minimal ${\ell }_1$-norm solution. Therefore, the locations of the faults can be obtained in a sequence, and only a consistently low computational load is required regardless of the isolated number of faults. The experimental results obtained with a scaled-down overactuated EV support the effectiveness of the proposed method, and a quantitative index of the sparsity condition for the target EV is discussed with a CarSim-connected MATLAB/Simulink simulation. Full article

Staphylococcus aureus is one of the most notorious Gram-positive bacteria with a very high mortality rate. The WHO has listed S. aureus as one of the ESKAPE pathogens requiring urgent research and development efforts to fight against it. Yet there is a major layback in the advancement of effective vaccines against this multidrug-resistant pathogen. SdrD and SdrE proteins are attractive immunogen candidates as they are conserved among all the strains and contribute specifically to bacterial adherence to the host cells. Furthermore, these proteins are predicted to be highly antigenic and essential for pathogen survival. Therefore, in this study, using the immunoinformatics approach, a novel vaccine candidate was constructed using highly immunogenic conserved T-cell and B-cell epitopes along with specific linkers, adjuvants, and consequently modeled for docking with human Toll-like receptor 2. Additionally, physicochemical properties, secondary structure, disulphide engineering, and population coverage analysis were also analyzed for the vaccine. The constructed vaccine showed good results of worldwide population coverage and a promising immune response. For evaluation of the stability of the vaccine-TLR-2 docked complex, a molecular dynamics simulation was performed. The constructed vaccine was subjected to in silico immune simulations by C-ImmSim and Immune simulation significantly provided high levels of immunoglobulins, T-helper cells, T-cytotoxic cells, and INF-γ. Lastly, upon cloning, the vaccine protein was reverse transcribed into a DNA sequence and cloned into a pET28a (+) vector to ensure translational potency and microbial expression. The overall results of the study showed that the designed novel chimeric vaccine can simultaneously elicit humoral and cell-mediated immune responses and is a reliable construct for subsequent in vivo and in vitro studies against the pathogen. Full article

For probing small distances in living cells, methods of super-resolution microscopy and molecular sensing are reported. A main requirement is low light exposure to maintain cell viability and to avoid photobleaching of relevant fluorophores. From this point of view, Structured Illumination Microscopy (SIM), Axial Tomography, Total Internal Reflection Fluorescence Microscopy (TIRFM) and often a combination of these methods are used. To show the high potential of these techniques, measurements on cell-substrate topology as well as on intracellular translocation of the glucose transporter GLUT4 are described. In addition, molecular parameters can be deduced from spectral data, fluorescence lifetimes or non-radiative energy transfer (FRET) between a donor and an acceptor molecule. As an example, FRET between the epidermal growth factor receptor (EGFR) and the growth factor receptor-bound protein 2 (Grb2) is described. Since this interaction, as well as further processes of cellular signaling (e.g., translocation of GLUT4) are sensitive to stimulation by pharmaceutical agents, methods (e.g., TIRFM) are transferred from a fluorescence microscope to a multi-well reader system for simultaneous detection of large cell populations. Full article