University of GuilanComputational Sciences and Engineering2783-25034220250419Analyzing bifurcation, stability and soliton solutions of Wang equation with a multiplicative white noise using Hamiltonian and Jacobian techniques187201868310.22124/cse.2025.30311.1102ENMostafaEslamiUniversity of MazandaranAnisEsmaeilyUniversity of MazandaranHamood UrRehmanDepartment of Mathematics, University of Okara, Okara, PakistanJournal Article20250209The nonlinear Schrödinger equation appears in many fields like quantum mechanics, optical fiber communications, plasma physics, and superfluid dynamics. In this context, we focused on the extended - dimensional stochastic NLSE. Specifically, we will explore these equations under the influence of multiplicative noise in the Itô framework. We apply the Sardar sub-equation method to investigate the exact solutions of the extended (3+1) - dimensional stochastic nonlinear Schrodinger equation under the influence of multiplicative noise. This method simplifies this nonlinear equation and derive the soliton-like, periodic, bright, dark and singular solutions, which are crucial for understanding wave propagation and stability in various physical systems. In this framework, bifurcation analysis allows us to explore how the system transitions at critical points or parameter thresholds. Chaotic behaviors are further examined by adding the external periodic functions. We can characterize regions where chaotic motion emerges, offering insights into unpredictable and turbulent behaviors that are common in plasma physics and optical fibers. Sensitivity analysis helps quantify how variations in system parameters influence the dynamics of the equation. By linearizing the system near equilibrium solutions, the stability of critical points is also investigated. Moreover, we present the behavior of these solutions graphically. By plotting the solutions obtained from the Sardar sub-equation method, we can observe the formation of solitons. Graphical illustrations of bifurcations, chaotic regimes and stability regions to enhance both qualitative and quantitative analysis of the system.https://cse.guilan.ac.ir/article_8683_9db0e384e97946357d3a348bfc0447b1.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220250623Advanced Deep Learning Approaches for Accurate and Efficient Suspicious Behavior Detection in Surveillance Videos203216884210.22124/cse.2025.30210.1099ENَArashSafdelFaculty of Engineering & Technology, University of Mazandaran, Babolsar, IranJamalGhasemiFaculty of Engineering & Technology, University of Mazandaran, Babolsar, IranSeyyed AliZendehbadFaculty of Engineering & Technology, University of Mazandaran, Babolsar, Iran0000-0002-5988-7043Journal Article20250327Violence Artificial Intelligence (AI) and Deep Learning (DL) systems present a difficult research area for identifying violence in videos within urban security frameworks and video surveillance systems. The proposed model divides violence detection tasks in video into two stages to achieve both rapid processing and precise outcomes. The LeNet-5 model operates at a speed of 0.8 frames per second to filter out non-violent videos during the first stage of operation. The second analysis stage employs the ResNet-50 model to inspect videos for potential violence when their probability surpasses 0.4. The Real-Life Violence dataset consisting of 1951 videos with 1000 violent and 951 non-violent videos was used for testing this system. The implementation produced 97.03% accuracy together with 95.70% recall and 98.46% precision and 97.06% F1-Score and AUC of 0.9902. Each frame requires only 20 milliseconds of processing time which allows real-time application of this system. A comparative analysis with existing methods, such as 3D-CNN, ViT, and YOLOv5+TSN, highlights the superiority of the proposed model in terms of both accuracy and speed. The system achieves better violence detection capabilities and operational reliability in real-world applications because it decreases detection errors.https://cse.guilan.ac.ir/article_8842_68fd9a2568497a504f4e0642705c06a7.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220240901An Integrated Multi-Objective MILP Model for Rebar Delivery Scheduling and Vehicle Routing: A Case Study217236892010.22124/cse.2025.31203.1113ENSeyed AminBadriUniversity of GuilanMehdiYaghoobiAmirkabir Khazar Steel Company, Lakan Industrial Town, Rasht, IranJournal Article20250315This study addresses the critical challenge of optimizing rebar delivery in heavy logistics industries by proposing an integrated multi-objective mixed-integer linear programming (MILP) model for simultaneous delivery scheduling and vehicle routing. The model aims to minimize three conflicting objectives: the overall makespan of deliveries, the weighted customer dissatisfaction from delivery time windows based on customer priority, and the total transportation costs. A fuzzy multi-objective optimization approach, based on the principles of Bellman and Zadeh and Zimmermann’s method, is employed to transform this complex problem into a single-objective maximization problem of an overall satisfaction level. The efficacy and practical applicability of the proposed model are validated through a real-world case study from Amir Kabir Khazar Steel Company in Gilan province, Iran. The case study involves 51 customer orders to be delivered over a three-day planning horizon, incorporating realistic constraints such as specific time windows and customer priority levels. Computational results, obtained using GAMS with the CPLEX solver, demonstrate that the model successfully achieves a high overall satisfaction level of λ =0.841. The findings offer significant managerial insights for balancing operational efficiency, cost reduction, and customer satisfaction in rebar supply chains.https://cse.guilan.ac.ir/article_8920_ad389038566df10b23df4398ca2aeb27.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220250703Efficient Pairwise Association Rules for Personalized Recommendations: Leveraging Caching and Asynchronous Model Updates237257887310.22124/cse.2025.30749.1108ENSeyed MohammadMortazaviAhrar Institute of Technology and Higher EducationFaridFeyziUniversity of GuilanJournal Article20250523Recommender systems based on content-based and collaborative filtering techniques face significant challenges, including the cold-start problem and privacy concerns due to their reliance on user profiles and product metadata. This study presents an optimized pairwise association rules (PAR) algorithm that addresses these limitations by operating independently of personal user data while maintaining recommendation accuracy. The proposed solution incorporates three key enhancements: (1) a privacy-preserving design using only transactional co-occurrence patterns, (2) a caching mechanism for modular training models that reduces recommendation latency by up to 102%, and (3) asynchronous execution for efficient resource management. Evaluations on a dataset of 20,000 food items demonstrate the algorithm's effectiveness, showing 18.7% higher nDCG scores than conventional methods while maintaining sub-second response times even with large-scale catalogs. The PAR algorithm proves particularly robust in sparse-data scenarios and cold-start conditions, offering a practical alternative to traditional approaches.https://cse.guilan.ac.ir/article_8873_eb161686c1c1ebfb89078bf4453c47f3.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220250625Nose-to-ID: A Deep Learning Framework for Dog Identification Using Nose-Print Biometrics259268884510.22124/cse.2025.30957.1110ENHediyehMosaferDepartment of Computer Engineering, Faculty of Technology and Engineering- East of Guilan, University of Guilan, Guilan, Iran.HomaTaherpour GelsefidDepartment of Computer Engineering, Faculty of Technology and Engineering- East of Guilan, University of Guilan, Guilan, IranRanaGhozatDepartment of Computer Engineering, Faculty of Technology and Engineering- East of Guilan, University of Guilan, Guilan, IranArminAzhdehniaDepartment of Computer Engineering, Faculty of Technology and Engineering- East of Guilan, University of Guilan, Guilan, IranJavadGhofraniDepartment of Computer Science, University of Applied Sciences Bonn-Rhein-Sieg, GermanyEhsanKozegarFaculty of Technology and Engineering, University of GuilanJournal Article20250606Accurate identification of individual dogs plays a crucial role in various applications including pet recovery, veterinary management, and animal welfare. This study proposes a fully automated dog identification framework based on unique nose-print biometric patterns, leveraging deep learning techniques to overcome limitations of traditional identification methods. The proposed approach processes user-submitted videos by selecting the most frontal frame via head pose estimation, detects the nose region using a fine-tuned YOLOv8 model, and extracts discriminative embeddings through a multi-resolution ResNeSt-based convolutional network enhanced with advanced augmentation strategies. The resulting embeddings are fused to produce robust identity descriptors capable of distinguishing between thousands of individual dogs. Experimental results demonstrate the system’s efficacy under real-world conditions, emphasizing its potential for practical deployment in pet identification and management systems.https://cse.guilan.ac.ir/article_8845_eb4c66b7807a3b563f49d660a11eb8df.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220250628Monitoring a Two-Stage Process Using Adaptive Control Charts with a Markov - Monte Carlo Chain Approach269282878610.22124/cse.2025.30676.1106ENKianoushFathi VajargahDepartment of Statistics, NT.C., Islamic Azad University, Tehran, Iran.HosseinEslami Mofid AbadiDepartment of Accounting & Management, Shahr.C., Islamic Azad University, Shahriar, Iran.Journal Article20250515One of the effective approaches to quality improvement is the application of statistical science within the framework of Total Quality Management (TQM). Statistical Process Control (SPC), as a key component of TQM, utilizes tools such as control sheets, histograms, Pareto charts, cause-and-effect diagrams, defect concentration charts, correlation diagrams, and control charts to detect and prevent defective products. This study focuses on control charts as instruments for identifying variations and out-of-control conditions in process means. In traditional methods, there is usually a delay between the occurrence of a process change and its detection on Shewhart control charts. This research aims to minimize such delay by proposing the use of adaptive control charts based on Markov chain models, which enhance the capability of rapid detection of assignable causes. To evaluate the proposed approach, one of the machines in a tea bag production company-characterized by a two-stage production process—was selected for case analysis. Sampling was conducted in two modes: once with fixed sample sizes and intervals, and again using adaptive sampling with variable sizes and intervals, to compare the efficiency of the proposed method.https://cse.guilan.ac.ir/article_8786_f59010fe50de630f7e60268df2707000.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220250816Improved Fuzzy Bayesian Reliability Analysis of Coherent Systems via the α-Pessimistic Method283296897910.22124/cse.2025.31257.1114ENMahnazMirzayiDepartment of Statistics, SR.C., Islamic Azad University, Tehran, IranRezaZareiDepartment of Statistics,
Faculty of Mathematical Sciences,
University of Guilan,
Rasht, Iran.GholamhosseinYariDepartment of Statistics, School of Mathematics, Iran University Science and Technology, Tehran, IranMohammad HassanBehzadiDepartment of Statistics, SR.C., Islamic Azad University, Tehran, IranJournal Article20250723In real-world reliability analysis, the underlying data and prior knowledge are often imprecise, posing significant challenges to classical probabilistic models. This study presents a novel fuzzy Bayesian approach for analyzing the reliability of coherent systems under imprecise prior information, where system lifetimes follow a Pascal distribution. We construct uncertain Bayes estimators using both squared error and precautionary loss functions by modelling the system reliability as a fuzzy random variable with a prior fuzzy distribution. A key innovation of the proposed approach is the application of the α-pessimistic method, which allows for the estimation process to be carried out without relying on complex non-linear programming, a common limitation in existing literature. Instead, this technique simplifies the computational procedure while enhancing interpretability and analytical tractability. The framework is applied to coherent systems, including parallel, series, and k-out-of-m structures, using Mellin transform techniques to derive the estimators. A numerical example is provided to demonstrate the practical applicability and effectiveness of the proposed method.https://cse.guilan.ac.ir/article_8979_3e06ed9751f8b2ee68101e8e4965782c.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220250906A New Insight on the Model of Support Vector Machine297308903710.22124/cse.2025.31583.1118ENAfsanehPourmoeziDepartment of Applied Mathematics, University of Mazandaran, Babolsar, IranMostafaEslamiDepartment of Applied Mathematics, University of Mazandaran, Babolsar, IranAliTavakoliDepartment of Applied Mathematics, University of Mazandaran, Babolsar, IranJournal Article20250703Support Vector Machine (SVM) is a powerful classification algorithm that separates samples by finding an optimal decision boundary. Its performance can degrade when feature variances differ across classes, potentially leading to suboptimal decision boundaries. A variance-weighted framework is proposed that reduces the influence of high-variance features while enhancing the impact of low-variance features, resulting in more accurate and robust decision boundaries. The method is applicable in both linear and nonlinear settings. Evaluation on synthetic datasets and real-world datasets, including Breast cancer and <em>a9a</em>, using cross-validation demonstrates that the variance-weighted SVM achieves higher accuracy and F1-score compared to soft SVM and LDM, particularly in scenarios with significant variance differences between classes.https://cse.guilan.ac.ir/article_9037_d305209df9180d00c147fe174517f0eb.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220240901On a best proximity point theorems for (α,d_G)-regular contractive type mappings in G-metric spaces309318919310.22124/cse.2024.26472.1071ENSami H.JasemDepartment of Mathematics, Faculty of Basic Sciences, Babol Noshirvani University of Technology, Babol, IranSayyed HashemRasouliDepartment of Mathematics, Faculty of Basic Sciences, Babol Noshirvani University of Technology, Babol, IranAzizollahBabakhaniDepartment of Mathematics, Faculty of Basic Sciences, Babol Noshirvani University of Technology, Babol, IranJournal Article20240114We consider the G-metric space with including the P-property which are introduced by Z. Mustafa and B. Sims (Nonlinear Convex Anal. 7 (2006) 289-297) and presented by B. Samet and et al. in a metric space (Nonlinear Anal. 4 (75) (2012) 2154-2165) respectively. In the present work we define P-property in a G-metric space and proved that under which various conditions there exist a best proximity point for non-self-mapping in G-metric space. Also, we introduced for a certain such mappings which its best proximity point is unique under further conditions.https://cse.guilan.ac.ir/article_9193_5ef772edb519e072ded860ede733d2b3.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220251119Analysis of a M[X]/G/1 queueing system with an unreliable server and delaying vacations using maximum entropy319330897210.22124/cse.2025.30255.1101ENMoeinMohsenzadehPh.D. student of Statistics, Faculty of Mathematical Sciences, Ferdowsi University of Mashhad, Mashhad, IranMahdiDoustparastDepartment of Statistics, Faculty of Mathematical Sciences, Ferdowsi University of Mashhad, Mashhad, IranAbdolrahimBadamchizadehDepartment of Statistics, Allameh Tabataba’i University, Tehran, Iran.MohammadJelodari MamaghaniDepartment of Mathematics, Allameh Tabataba'i University, Tehran, IranJournal Article20250402This paper deals with a single unreliable server and with delaying<strong> </strong>vacations which has Poisson arrivals and general distribution for the service times. The server can be activated at arrival epochs or deactivated at service completion epochs. The maximum entropy principle is increasingly relevant to queueing systems. The principle of maximum entropy (PME) presents an impartial framework as a promising method to examine complex queuing processes. We use maximum entropy principle to derive the approximate formulas for the steady-state probability distributions of the queue length. The maximum entropy approach is then used to give a comparative perusal between the system’s exact and estimated waiting times. We demonstrate that the maximum entropy approach is efficient enough for practical purpose and is a feasible method for approximating the solution of complex queueing systems.https://cse.guilan.ac.ir/article_8972_df034b0035a1de40b00b8342f51c5b17.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220251205Monte Carlo simulation in acceleration Kaczmarz method by the Johnson–Lindenstrauss lemma331349911010.22124/cse.2025.31590.1119ENSomayehAghaei KhomamiRasht MunicipalityJournal Article20250804In this paper, we propose an accelerated variant of the randomized Kaczmarz method for solving large-scale linear systems, including both standard and inequality-constrained systems. The key innovation lies in integrating the Johnson–Lindenstrauss (JL) lemma into the row-selection process, which allows high-dimensional rows to be projected onto lower-dimensional spaces while approximately preserving pairwise distances. This enables near-optimal row selection with reduced computational cost, improving both convergence rate and stability, particularly for ill-conditioned systems. Furthermore, Monte Carlo techniques are employed to efficiently construct the projection matrices, enhancing the overall computational performance. Numerical experiments demonstrate that the proposed method achieves faster convergence and higher accuracy compared to traditional randomized Kaczmarz and other conventional techniques, making it highly suitable for large-scale problems in applied mathematics and engineering.https://cse.guilan.ac.ir/article_9110_71727176d5e5cafe414a8f241347260d.pdfUniversity of GuilanComputational Sciences and Engineering2783-25034220240901The k-order variations on the k-Fibonacci universal code for k ≥ 4351360916710.22124/cse.2025.31829.1126ENMansour HashemiBaragooriFaculty of mathematical sciences, University of guilan, IranJournal Article20250830In the current paper, first we study the <em>k−</em>order variations on the <em>k</em>-Fibonacci universal code for <em>k ≥</em> 4 (denoted by ) and get Table 3. Also in Tables 1 and 2, we obtain the <em>k</em>-Fibonacci representation and the <em>k</em>-Fibonacci code for <em>k ≥</em> 4 and 1 <em>≤ n ≤</em> 50, respectively. Finally, we examine a new blocking algorithm using <em>k−</em>Fibonacci code and Gopala-Hemachandra code.https://cse.guilan.ac.ir/article_9167_9878563d55aefeba44437315d47a1a13.pdf