This paper applies unsupervised data mining to segment users of information-library centers by their perception of digital and marketing communication. The empirical basis is a survey of readers from fifteen information-library centers in Uzbekistan. From 213 responses, 180 complete records with seven Likertscale perception variables were retained. After z-score standardization, k-means clustering was used, and the number of clusters was selected by the silhouette coefficient with additional validation by the CalinskiHarabasz and Davies-Bouldin indices. A stable four-cluster solution was obtained: Engaged Enthusiasts, Moderately Satisfied, Loyal-in-spite-of and Disengaged. The main finding is a perception-loyalty paradox: some users remain satisfied even when mediated communication is weakly perceived. The proposed segmentation provides a reproducible instrument for differentiated communication planning in library management.
The article discusses the problem of diagnosing diseases of crops based on machine learning methods. A hybrid GA-SVM algorithm is proposed, combining the support vector machine (SVM) method and a genetic algorithm for optimizing the hyperparameters of the model. A comparative analysis with the classical GridSearch-SVM method is carried out. Experimental results have shown that the proposed approach provides higher classification accuracy and model stability. Improved Accuracy, Precision, Recall and F1-score indicators have been achieved, as well as a reduction in the number of classification errors. Despite the increased computational costs, the GA-SVM method demonstrates a more efficient search for optimal parameters. The results obtained confirm the prospects of using hybrid algorithms to improve the quality of diagnosis of plant diseases and can be applied in intelligent agricultural systems.
This article examines the conceptual model, architecture, and structure of the information system for paraphrasing texts in the Uzbek language. The conceptual model represents all processes in the system. In addition, the processes of paraphrasing sentences in the Uzbek language are also separately expressed. The purposes of use of the information system users are separately indicated. The developed information system also has the ability to identify paraphrased sentences in documents, which makes it easier to check publications in editorial offices, identify and isolate duplicate sentences in scientific articles and dissertations. The general architecture and user interface of the system have been developed. The roles of user, specialist, and programmer have been introduced in the system, and separate functions have been defined and analyzed for each role.
Modern Internet traffic is dominated by a few global platforms, making accurate service identification essential for QoS management and network analytics. Yet pervasive encryption and shared infrastructure increasingly limit port-based methods and DPI. This paper presents a two-stage hierarchical framework for classifying encrypted Google-family traffic (YouTube, Gmail, Google Search) using only flow-level statistical and temporal features extracted with CICFlowMeter. Information Gain is applied to reduce feature redundancy and computational cost. As a baseline, a single-stage Random Forest achieved an overall accuracy of 0.89 but showed strong confusion between Gmail and Google Search. In the proposed framework, Stage 1 separates YouTube vs Other, and only flows predicted as Other are forwarded to Stage 2, where Gmail vs Search are distinguished. This staged design increased the overall accuracy from 0.89 to 0.96, with ROC-AUC of 0.99 for Stage 1 and 0.93 for Stage 2. The results indicate that hierarchical classification effectively mitigates service ambiguity in encrypted, shared-infrastructure environments and supports scalable telecom monitoring.
Auditing encryption and key management policies in modern web and server systems is complicated by architectural complexity and continuous configuration change. Existing approaches largely rely on static compliance checks or isolated metrics, providing limited support for actionable decision-making. This paper proposes a decision-oriented framework that bridges metric-based auditing and practical security governance. The framework relies on system-level abstractions of policy requirements and enforcement evidence, and maps consistency, conflict, stability, and risk metrics to discrete decision outcomes. A bounded and non-intrusive satisfaction function supports partial compliance, heterogeneous evidence, and conservative handling of missing data without accessing cryptographic key material. In addition, a riskaware remediation prioritization algorithm ranks policy requirements by urgency and architectural impact. Scenario-based evaluation demonstrates improved interpretability of audit results and supports proactive, risk-aware remediation planning.
Hyperspectral images contain a huge number of spectral channels, which ensures high accuracy of analysis, but at the same time leads to problems associated with data redundancy, high computational load and decreased classification efficiency. In this paper, a method is proposed to integrate time series complexity analysis and fractal dimension (FD) to effectively reduce the dimensionality of hyperspectral data. Each pixel is considered as a time series characterized by spectral complexity, which allows identifying the most informative parts of the spectrum. Fractal dimension is used to quantify the complexity of spectral features and select significant channels. The proposed approach allows preserving critical information while minimizing losses during dimensionality reduction. To evaluate the effectiveness of the method, a comparison of the classification accuracy using the support vector machine (SVM) algorithm was carried out before and after applying the proposed optimization procedure. Experimental results on real hyperspectral data show significant dimensionality reduction while maintaining or improving classification quality.
Achieving high predictive accuracy while utilizing a minimal yet highly relevant set of features remains a critical challenge in machine learning. To address this, numerous feature selection techniques—ranging from ANOVA and LASSO to supervised methods like Mutual Information—have been developed, each offering unique strengths and limitations. Building upon these foundations, we introduce innovative hybrid approaches that seamlessly integrate filter and wrapper methods for more effective feature selection. These approaches were rigorously tested across multiple models. In this research, the new approaches of the integration of Chi-square test (with new step: relationship level) with SFS algorithm are proposed for enhancement model performance in feature selection. By the help of the new approaches, the trained results were 89% (in SVM), 88% (in kNN), 88% (in RF), 91% (in FCNN). These outputs are better than other feature selection methods’ results.
This study proposes a semantically enriched multimodal neural architecture for automatic translation of Uzbek Sign Language (UZSL). In the approach, first, the text stream is brought to semantic consistency through n-gram tokenization, lemmatization, and morphological normalization, and a high-information context is prepared for BERT-MLM. Then, hand, face, and body landmarks are extracted from video frames using MediaPipe Holistic, speech transcriptions are generated using Speech-to-Text, and they are annotated with gloss dictionary-based labels in the text→gloss→landmark relationship. During the translation process, the text is fed to BERT-MLM, and semantic candidates are generated using a Top-k masking strategy; these candidates are checked against a standard gloss corpus in the constraint layer and filtered using semantic similarity calculations based on Word2Vec. Finally, a re-ranking mechanism based on SBERT evaluates the cosine similarity between the context and glosses and selects the most appropriate gloss. This integrated pipeline provides a near-real-time, consistent, and scalable translation workflow for UZSL by reducing the morphological complexity of the text, extracting stable features from multimodal signals, and selecting glosses based on semantic criteria with dictionary constraints.
In this research, the wear dynamics of the experimental working body named “Soil-Tillage Composite Working Organ,” patented as utility model No. FAP 02305 by the Ministry of Justice of the Republic of Uzbekistan, were investigated under field conditions. The experimental design aimed to compare the wear resistance of the new composite working organ with the conventional arrow-shaped tines used in KXU-4M cultivator units. Field tests were conducted in the Jizzakh region on cotton inter-row cultivation areas, and measurements were systematically recorded during soil treatment processes covering up to 76.5 hectares. The geometric wear parameters, including coverage width, blade thickness, and sharpening angle, were analyzed. Statistical analysis of the results showed that the newly designed working organ exhibited durability 1.8–2.0 times greater than that of the conventional arrow-shaped tine. The proposed design eliminates the main shortcomings of traditional tines by using double parallelogram-shaped modules that can be rotated and reused, ensuring longer operational life and stable agro-technical performance.
Accurate crop yield prediction is critical for enhancing food security, particularly in agrarian economies prone to soil degradation and climatic uncertainties. This study explores the application of Support Vector Regression (SVR) for forecasting wheat yields in Uzbekistan, utilizing soil fertility indicators as key predictive features. Unlike conventional linear regression models, SVR effectively captures complex nonlinear interactions between soil physicochemical properties and crop productivity, thereby offering improved adaptability to real-world agricultural conditions. The dataset comprises essential soil attributes, including nitrogen (N), phosphorus (P), potassium (K), pH, organic carbon (OC), electrical conductivity (EC), and micro-nutrient concentrations. Data preprocessing involved feature standardization, K-nearest neighbor (KNN) imputation for handling missing values, and correlation analysis to select the most influential variables. The dataset was partitioned using an 80/20 stratified split, and the SVR model with a radial basis function (RBF) kernel was optimized through 5-fold crossvalidation and exhaustive grid search for hyperparameter tuning. The optimized SVR model achieved a coefficient of determination (𝑅 2 ) of 0.87 and demonstrated a low root mean square error (RMSE), outperforming baseline regression methods. Model interpretability was enhanced using SHapley Additive exPlanations (SHAP), which identified soil pH, organic carbon, and available phosphorus as the most significant predictors of wheat yield—findings consistent with established agronomic principles. Overall, the results confirm SVR’s potential as a robust, scalable, and interpretable tool for precision agriculture, offering practical insights for site-specific yield forecasting and promoting sustainable land management practices in Uzbekistan.
This article examines the aspects that influence productivity in distant service systems during times of request congestion. At the same time, the issue of enhancing system server efficiency without the use of additional hardware or software has been investigated. It has been found that service system efficiency declines during traffic congestion because requests are processed in multiple stages, each with a different service order. Hence, to enhance system efficiency during traffic congestion, a proposal is made to prioritize users whose requests have been successfully served at least once. A physical model of the service process was constructed to assess the efficacy of the proposed approach.The research conducted using this model contributed to an average increase in the efficiency of service systems by 6%.
In the article, the bending of the cultivator paws interaction of variables in the study issues of determining statistical relationships studied. It's razor sharp with the size of the case and the statistic between the back chamfer corners very weak correlation (r0.0403 and r0.0632), i.e. found to be non-linear. The width of the back bevel with the size of the work, the blade of the razor according to the width and thickness of the blade and statistical correlations between bending the higher the correlation coefficients values range from r0.91 to 0.96. Work many parameters or factors in the developed program also in studying the connections between can be used successfully emphasized.
Advancements that merge the clarity of symbolic AI with the adaptive learning traits of subsymbolic AI show great potential at the intersection of these two AI forms. This study introduces Fuzzy Cognitive Maps (FCMs). This hybrid model integrates the optimal characteristics of both frameworks to address the challenges of interpretability and explainability in artificial intelligence (AI) systems. FCMs provide a robust framework for logically and intuitively supporting decision-making processes and representing causal relationships. Their capacity to handle the inherent vagueness and uncertainty of real-world scenarios enables a more natural and flexible approach to problem-solving. Due to their intrinsic adaptability and learning capabilities derived from sub-symbolic AI, FCMs are particularly suited for applications demanding high levels of interpretability and explainability.
In this paper is given an overview of tools and techniques for obtain information about the geometry of 3D scenes from 2D images. The generating three-dimensional structure from a series of 2D images or video of scenes is known as Structure from Motion (SfM). A central tenet of structure from motion is that, given the position of a feature in one image, it is possible to find the corresponding position of the same feature in successive images. We described methods for the simultaneous recovery of 3D points and camera projection matrices using corresponding image points in multiple views. Structure from Motion (SfM) is a fascinating field within computer vision that seeks to reconstruct a three-dimensional structure of the environment from a sequence of two-dimensional images. At the heart of SfM lies a set of sophisticated algorithms that enable the extraction of spatial information from a series of images. Two public repositories that may be of use are Open SfM and Colmap.
Recent decades have witnessed a significant increase in the use of visual odometry(VO) in the computer vision area. It has also been used in varieties of robotic applications, for example on the Mars Exploration Rovers. This paper, firstly, discusses two popular existing visual odometry approaches, namely LSD-SLAM and ORB-SLAM2 to improve the performance metrics of visual SLAM systems using Umeyama Method. We carefully evaluate the methods referred to above on three different well-known KITTI datasets, EuRoC MAV dataset, and TUM RGB-D dataset to obtain the best results and graphically compare the results to evaluation metrics from different visual odometry approaches. Secondly, we propose an approach running in real-time with a stereo camera, which combines an existing feature-based (indirect) method and an existing feature-less (direct) method matching with accurate semidense direct image alignment and reconstructing an accurate 3D environment directly on pixels that have image gradient. Keywords VO, performance metrics, Umeyama Method, feature-based method, feature-less method & semi-dense real-time.
In this paper, we present a novel solution to detect forgery and fabrication in passports and visas using cryptography and QR codes. The solution requires that the passport and visa issuing authorities obtain a cryptographic key pair and publish their public key on their website. Further they are required to encrypt the passport or visa information with their private key, encode the ciphertext in a QR code and print it on the passport or visa they issue to the applicant. The issuing authorities are also required to create a mobile or desktop QR code scanning app and place it for download on their website or Google Play Store and iPhone App Store. Any individual or immigration uthority that needs to check the passport or visa for forgery and fabrication can scan its QR code, which will decrypt the ciphertext encoded in the QR code using the public key stored in the app memory and displays the passport or visa information on the app screen. The details on the app screen can be compared with the actual details printed on the passport or visa. Any mismatch between the two is a clear indication of forgery or fabrication. Discussed the need for a universal desktop and mobile app that can be used by immigration authorities and consulates all over the world to enable fast checking of passports and visas at ports of entry for forgery and fabrication
Our research focuses in software-intensive organizations and highlights the challenges that surface as a result of the transitioning process of highly-structured to DevOps practices and principles adoption. The approach collected data via a series of thirty (30) interviews, with practitioners from the EMEA region (Czech Republic, Estonia, Italy, Georgia, Greece, The Netherlands, Saudi Arabia, South Africa, UAE, UK), working in nine (9) different industry domains and ten (10) different countries. A set of agile, lean and DevOps practices and principles were identified, which organizations select as part of DevOps-oriented adoption. The most frequently adopted ITIL® service management practices, contributing to DevOps practice and principle adoption success, indicate that DevOps-oriented organizations benefit from the existence of change management, release and deployment management, service level management, incident management and service catalog management. We also uncover that the DevOps adoption leadership role is required in a DevOps team setting and that it should, initially, be an individual role.
AI-based security systems utilize big data and powerful machine learning algorithms to automate the security management task. The case study methodology is used to examine the effectiveness of AI-enabled security solutions. The result shows that compared with the signature-based system, AI-supported security applications are efficient, accurate, and reliable. This is because the systems are capable of reviewing and correlating large volumes of data to facilitate the detection and response to threats.
A well-constructed classification model highly depends on input feature subsets from a dataset, which may contain redundant, irrelevant, or noisy features.This challenge can be worse while dealing with medical datasets.The main aim of feature selection as a pre-processing task is to eliminate these features and select the most effective ones.In the literature, metaheuristic algorithms show a successful performance to find optimal feature subsets.In this paper, two binary metaheuristic algorithms named S-shaped binary Sine Cosine Algorithm (SBSCA) and V-shaped binary Sine Cosine Algorithm (VBSCA) are proposed for feature selection from the medical data.In these algorithms, the search space remains continuous, while a binary position vector is generated by two transfer functions S-shaped and V-shaped for each solution.The proposed algorithms are compared with four latest binary optimization algorithms over five medical datasets from the UCI repository.The experimental results confirm that using both bSCA variants enhance the accuracy of classification on these medical datasets compared to four other algorithms.
The purpose of this paper is to investigate the e-commerce credibility factors affecting the perception of users in Saudi Arabia and, moreover, to investigate whether the variation of credibility factors in Saudi Arabian e-commerce websites influence users' performance. Website credibility, which refers to the believability of the website and its content, plays an important role in consumers’ successful online shopping experience and satisfaction. This investigation is conducted by employing two credibility evaluation methods: heuristic evaluation and performance measurement. This study adopts Fogg's 10 Stanford credibility guidelines as a starting point for the heuristic evaluation. In the performance measurement method, two measurements are used: the amount of time needed to finish the task and the total number of clicks taken to finish the task. A frequency analysis of the comments and a one-way ANOVA test are used to establish the results. Three e-commerce websites in Saudi Arabia are selected. The findings show that Fogg’s 10 Stanford credibility guidelines can be implemented in the Saudi Arabian e-commerce context with minor modifications and expansions by adding reputation, endorsement, security, and service diversity guidelines. Another important finding is that professional website design plays a vital role in users' first impression of websites, while usability is the most important credibility factor investigated used to evaluate the credulity of an e-commerce website. Lastly, the results of this study indicate a relationship between the e-commerce credibility level and users’ performance. This paper contributes to the literature by providing a set of credibility guidelines associated with specific criteria, which can be assessed to improve the future of e-commerce in Saudi Arabia.
