Vagif Gasimov

Vagif Gasimov

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Vagif Gasimov has a total of 22 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    MACHINE LEARNING APPLICATION FOR INTELLIGENT THREAT CLASSIFICATION IN CLOUDFLARE BOT MANAGEMENT FIREWALL LOGS OF UNIVERSITY DIGITAL EDUCATION INFRASTRUCTURE

    by Artughrul Gayibov, Vagif Gasimov, Islam Islamov, Jabir Mammadov, Seyran Gayibov

    The object of this study is the automated classification of Cloudflare Bot Management (CBM) firewall log records collected from a university web application. The study addresses the problem of distinguishing potentially malicious requests from legitimate academic traffic in the digital educational infrastructure of a higher education institution. Static threshold-based firewall rules are often insufficient for analyzing multidimensional security indicators, including WAF machine-learning attack scores, bot-related signals, TLS fingerprint information, and request-level metadata. A comparative machine-learning framework was developed using six supervised classifiers: Random Forest (RF), Gradient Boosting (GB), Decision Tree (DT), Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Gaussian Naive Bayes (GNB). The models were trained and evaluated on 24 engineered features extracted from 500 CBM log entries using a stratified 80/20 train-test split and 5-fold cross-validation. RF, GB, and DT achieved 100% test accuracy on the constructed dataset, while SVM, KNN, and GNB achieved 96.00%, 93.00%, and 44.00%, respectively. These results should be interpreted in the context of the proposed labelling procedure, because the target variable was derived from WAF score aggregation and the same group of WAF indicators was also used among the input features. Feature importance analysis showed that WAF machine-learning attack scores formed the dominant group of classification indicators. The Decision Tree model produced an interpretable threshold rule based on the normalized “wafMlAttackScore”. This rule is not presented as a fully deployed production firewall policy, but as a candidate rule that can support further expert validation and configuration of university WAF security policies. The proposed approach therefore contributes to the interpretation and secondary classification of CBM firewall logs in educational ICT security management.

  • Open Access

    Linear-Time Correlation-Controlled Shuffling on Random Regular Graphs with Spectral Guarantees

    by Vagif Gasimov, Nargiz Mammadzada, Esmira Mustafayeva, Jabir Mammadov, Kamala Alieva

    This article proposes a linear-time correlation-controlled shuffling method that attenuates Pearson correlation by permuting one variable through local swaps on a random regular graph. Each round processes a fixed edge order and accepts only vertex-disjoint swaps, producing greedy matching rather than a standard interchange-process update. The analysis uses the induced first-moment operator. A uniform lower bound on edge acceptance yields a Laplacian-domination relation, while a lower bound on the probability that vertices remain unmatched controls the negative spectrum. Conditional on the loop-deleted realized graph being connected and having a spectral gap bounded away from zero, the absolute value of the expected correlation contracts exponentially. For every fixed same-sign attenuation target, an activation probability exists that attains the target in expectation. Finite-sample calibration uses a fixed-budget multi-resolution direct search without assuming monotonicity. Because the procedure applies only permutations, the marginal distribution is preserved exactly. Under fixed degree and fixed calibration budgets, both calibration and shuffling scale linearly with sample size. Experiments show small target errors, exact marginal preservation, and approximately linear runtime scaling. A matched-degree spectral ablation further shows that, with degree and edge count held fixed, the higher-gap random regular graph exhibits substantially faster decay of the mean correlation than the low-gap regular circulant graph.

  • Open Access

    Evaluation of a machine learning-assisted interactive evolutionary non-dominated sorting genetic algorithm -II framework for hyperparameter optimization of U-Net in agricultural land segmentation

    by Artughrul Gayibov, Vagif Gasimov, Esmira Mustafayeva, Kamala Aliyeva, Dilara Guluzada

    The object of the study is the hyperparameter configuration space of the U-Net architecture for agricultural land segmentation from Sentinel-2 satellite imagery. The problem being solved is the excessive cost of multi-objective hyperparameter optimization, because non-dominated sorting in the non-dominated sorting genetic algorithm II (NSGA-II), with complexity O(MN2), becomes a bottleneck for deep segmentation models. To address this problem, an interactive evolutionary non-dominated sorting genetic algorithm II (IENSGA-II) framework is evaluated, in which a logistic regression classifier is trained on hyperparameter vectors and Pareto ranks from initial NSGA-II generations, then used to predict ranks in subsequent generations instead of full sorting. Unlike existing surrogate-assisted approaches, this work predicts Pareto ranks without additional model evaluations. On the panoptic agricultural satellite time series (PASTIS) benchmark, the framework reduced execution time by 20.07%, 16.39%, and 38.80% for 5, 10, and 15 generations, and in the 10-generation setting improved validation criteria, reaching an area under the receiver operating characteristic curve (AUC) of 0.9072 versus 0.9004 and validation loss of 0.6057 versus 0.6212. These results were achieved because the method accelerates selection rather than replacing model evaluation, while AUC-based tie-breaking preserves preference for more accurate solutions among candidates with same predicted rank. Effectiveness stems from a regular relationship between hyperparameters and Pareto ranks in early evolutionary data. In practice, the method is used in resource-constrained multi-objective learning when initial generations provide representative data for rank prediction

  • Extending spectral indices from multispectral satellite data using U-Net segmentation

    by Artughrul Gayibov, Vagif Gasimov

    This study focuses on developing a unified, reproducible cloud-to-model pipeline for parcel-scale cropland delineation from multispectral Sentinel-2 and Landsat imagery in the Kur–Araz region. The goal of this study is to produce accurate, boundary-faithful, and computationally efficient farmland maps that remain transparent, scalable, and deployable on commodity hardware for use by government agencies, water authorities, and agricultural producers. The tasks to be addressed include the following: specification of the study scope, data sources, and evaluation protocol that integrates pixel-wise and boundary-sensitive accuracy metrics; construction of a multi-index feature stack in addition to surface reflectance bands, followed by screening of features for cross-seasonal stability; design and training of a memory-efficient U-Net architecture with a hybrid loss that simultaneously balances calibration and overlap; and validation of model generalization across different cropping seasons and neighboring subregions of Kur–Araz, with complete provenance tracked from preprocessing through evaluation. The methods used in the study include cloud-native preprocessing in Google Earth Engine, consisting of cloud masking, seasonal compositing, medoid and percentile mosaicking, and stratified patch sampling with spatial blocking. The datasets were exported to TFRecords and used to train a compact U-Net encoder–decoder network with skip connections and a hybrid objective function. Training involved strong normalization, data augmentation to account for agricultural scene variability, overlap-tiling for inference, and three ablation studies to isolate the contribution of individual vegetation indices. Complete logging of assets, random seeds, and parameters supported exact reruns of the pipeline, replication, and auditing. Conclusions. The scientific novelty of the results obtained is as follows: (1) a unified, provenance-complete pipeline was developed that directly integrates Google Earth Engine preprocessing with deep learning model training and boundary-aware evaluation, ensuring full reproducibility and auditability; (2) a resource-efficient U-Net architecture with skip connections and a hybrid overlap-weighted loss was designed and validated, which preserves parcel borders while maintaining calibration, outperforming heavier transformer-based backbones under limited compute budgets; (3) the complementarity of multi-index feature stacks was systematically quantified through ablation studies, demonstrating their role in mitigating spectral saturation and soil brightness effects and improving overlap fidelity. Finally, a standardized evaluation protocol was proposed that combines pixel-wise and boundary-sensitive metrics, allowing robust assessment of farmland delineation performance across seasons and spatial subregions.

  • Interactive Evolutionary NSGA-II with Machine Learning for Efficient Hyperparameter Optimization of Convolutional Neural Networks

    by Artughrul Gayibov, Vagif Gasimov

    This paper introduces a hybrid optimization framework that integrates a Machine Learning (ML) model with the Non-Dominated Sorting Genetic Algorithm-II (NSGAII) for efficient hyperparameter tuning of Convolutional Neural Networks (CNNs). The application focuses on the MNIST dataset for handwritten digit recognition. The proposed Interactive Evolutionary NSGA-II (IE-NSGA-II) leverages a Decision Tree classifier to predict Pareto front ranks after an initial phase of traditional NSGA-II execution. This approach aims to reduce computational overhead associated with standard NSGA-II procedures without significantly compromising solution quality. Experimental results demonstrate that the IE-NSGA-II achieves competitive performance compared to the traditional NSGA-II. At 20 generations, the IE-NSGA-II obtained a validation accuracy of 99.02 % with a validation loss of 0.0349, while the traditional NSGA-II achieved 99.08 % accuracy and 0.0335 loss. The IE-NSGA-II showed a modest reduction in mean execution time, being approximately 4.63 % faster than traditional NSGA-II at 20 generations. This study highlights the potential of MLassisted evolutionary algorithms for creating scalable and efficient hyperparameter optimization techniques in deep learning.


  • Open Access

    The problem of information protection in electronic document management systems of medical organizations

    by V.A. Gasimov, N.F. Mammadzada, J.I. Mammadov, K.J. Aliyeva

    The article is devoted to the study and processing of methods for creating electronic document management systems of medical organizations, increasing the level of protection of documents and information stored and processed in this system. For this purpose, International and local experience in the field of legislation on the protection of documents containing personal and medical information of patients in medical organizations was commented in the article, the issue of creating a distributed structured electronic document management system for medical organizations was considered, various protection methods were used to ensure the confidentiality of personal and medical information of patients, it was proposed to use the authentication procedure and secure encryption methods.

  • Open Access

    ANALYSIS OF THE RESULTS OF SIMULATIVE MODELLING OF THE INFORMATION SECURITY SYSTEM IN THE CORPORATE NETWORKS OF HIGHER EDUCATION INSTITUTIONS

    by Vagif Gasimov, Balami Ismailov

    The analysis shows that the insufficient level of information security in service networks is the main cause of huge losses for enterprises. Despite the appearance of a number of works to solve this problem, there is currently no unified system for assessing information security. This shows that this problem has not yet been sufficiently studied and relevant. This work is one of the steps towards creating a system for assessing information security in service networks. The purpose of the work is to develop an algorithm and simulation model, analyze the results of the simulation model to determine the main characteristics of the information security system (ISS), providing the ability to completely close all possible channels of threats by controlling all unauthorized access (UA) requests through the protection mechanism (PM). To solve the problem, a simulation method was applied using the principles of queuing systems (QS). This method makes it possible to obtain the main characteristics of the ISS from the UA with an unlimited amount of buffer memory (BM). Models, an algorithm and a methodology for the development of ISS from UA are proposed, which is considered as a single-phase multi-channel QS with an unlimited volume of BM. The process of obtaining simulation results was implemented in the GPSS World modeling system and comparative analyzes of the main characteristics of the ISS were carried out for various laws of distribution of output parameters. At the same time, UA requests were the simplest flows, and the service time was subject to exponential, constant and Erlang distribution laws. Conducted experiments based on the proposed models and algorithm for analyzing the characteristics of the ISS from the UA as a single-phase multi-channel QS with unlimited waiting time for requests in the queue confirmed the expected results. The results obtained can be used to build new or modify existing ISS in corporate networks for servicing objects for various purposes. This work is one of the approaches to generalizing the problems under consideration for systems with an unlimited volume of BM. Prospects for further research include research and development of the principles of hardware and software implementation of ISS in service networks.

  • Maze based image encryption method constructed by random number generation

    by Azerbaijan Technical University, V Gasimov, N Mammadzada, J Mammadov, E Mustafayeva

    Due to the large size of image files and the high correlation between the numerical values of the pixels, various improved algorithms have been used for image encryption in recent years. In most cases improvement is obtained by taking into account the characteristics of chaotic and random processes in encryption algorithms. In this study, an algorithm is proposed that encrypts images based on a maze created by keys consisting of random numbers generated with Feigembaum’s quadratic function. For the generation of the maze, deep first search (DFS) algorithm has been used. During the generation with this algorithm, numbers representing the walking paths and as well as the length of the used stack are included in the encryption process. Algorithm security verification was carried out according to key space and differential analysis, statistical analysis such as entropy estimation, and analysis of correlation between the histogram and neighboring pixels of the image.

  • Open Access

    Data Management as a Critical Component of Protecting Corporate Devices

    by Agassi Melikov, Vagif Gasimov, Samur Ahmadov

    Data Management as a Critical Component of Protecting Corporate Devices


  • Open Access

    Implementation of industrial internet of things technologies using Edge, Fog, and Cloud computing

    by V.A. Gasimov, Sh.Kh. Aliyeva, Zh.S. Assanova

    Having superior features of productivity of the Industrial Internet technology of things used in the establishment of clever cities and clever villages, optimity, etc. increases the need to use it. The low-capacity of the physical resources of Iot-based sensors allows it to integrate into cloud technology. Cloud-based infrastructures are not sufficient for the existing requirements of the products of the products. The two main issues, in terms of reconciliation and network expansion limits arise. Also, the breaks of the Internet connection, the low conductivity, etc., as such problems, allow IIOT to integrate the "Fog Computing" and "Edge Computing" technologies. Taking into account the coverage and development of IIOT technology, it is believed that in many cases taking advantage of its full potential will be possible to combine clouds, fog and edge calculations. In this article, Industrial Internet technology of things, its integrated "Cloud", Fog ", Edge technologies are being investigated and differences between them, distribution-based approaches between levels are discussed.