Informatics and cybernetics

Research on machine and deep learning algorithms for detecting tumors in the human brain

2025-01-14T13:45:00+00:00

Modern medical imaging research faces the challenge of identifying brain tumors using magnetic resonance imaging (MRI). A brain tumor is an abnormal mass of tissue in which some cells multiply uncontrollably, apparently unregulated by the mechanisms that control normal cells. There are three types of tumors that are commonly observed, namely: benign, precancerous and malignant. Many supervised and unsupervised methods are classified to determine a tumor as benign or malignant. Generally, lighter weight datasets are used for image classification in the application domain, while heavier and heavier datasets are used in the medical domain. Many parameters selected during training play a critical role in the performance and accuracy of the systems. Thus, an attempt was made to visualize how the accuracy of the algorithm increases depending on the parameters chosen to detect the human brain in an MRI image.

Expanding the capabilities of image generation systems by using neural networks

2025-01-14T14:03:25+00:00

The article provides an overview of the main challenges faced by researchers and practitioners in generating images using neural networks. Key aspects such as computational complexity, power consumption and quality of the generated images are discussed. The paper also suggests potential solutions to these problems, including optimization of neural network architectures, application of optimization techniques, and use of specialized hardware gas pedals. The prospects for research in this area are summarized, and directions for future research and innovation are outlined.

Intellectualization algorithms and their application in automated control systems

2025-01-14T14:16:47+00:00

An algorithm is proposed for finding the best solutions when improving the operation of the software part of ACS (automated control systems), which allows optimizing the expenditure of ACS resources to perform various tasks. The problems of optimizing algorithms for a specific task are analyzed. The algorithm for finding the best solutions also takes into account the time spent and has specific measured results, which makes it possible to draw intermediate conclusions about its effectiveness. The results of research are done.

Investments in artificial intelligence: assessment of economic feasibility and potential risks

2025-01-14T14:27:17+00:00

The article evaluates the economic feasibility of investing in artificial intelligence and identifies potential risks. The prospects for the growth of the AI market, the benefits of its implementation and methods for evaluating profitability are discussed. Technical, legal and ethical risks are analyzed, as well as methods to reduce them. Proper assessment of benefits and risks will help you make informed decisions on investing in artificial intelligence and achieve optimal results.

Application of convolutional neural networks for object recognition in an image.

2025-01-14T14:58:12+00:00

The analysis of the principles of artificial neural networks, the scope of their application for recognizing objects in the image is carried out. An algorithm based on convolutional neural networks has been selected and implemented, capable of detecting selected classes of objects in an image with fairly high accuracy. Using this algorithm to analyze video streams received from self-service ticket offices can allow an enterprise, such as a cafe hall or a fast food restaurant, to function without human intervention.

Text processing using natural language methods

2025-01-14T15:06:40+00:00

The paper examines text processing methods using natural language (NLP), which plays a key role in the modern world of information technology. The article covers basic NLP concepts and techniques such as tokenization, stemming, lemmatization, stop word removal, regular expressions, and text representation methods including Bag of Words and TF-IDF. Particular attention is paid to sentiment analysis, machine translation, automatic summarization and chatbots, which are important areas in the field of NLP.

Video recording of students' presence in the classroom based on neural network facial recognition.

2025-01-14T15:18:42+00:00

This paper describes the architecture of the convolutional neural network VGGFace, on the basis of which the scheme of computer recognition of a person by his face is developed. The program realization of the system, which performs automatic maintenance of the electronic group journal, is performed. Experiments on video registration of students at the entrance to the classroom with the help of computer vision are carried out.

Rational approaches to the implementation digital twins in water supply and sewerage systems.

2025-01-14T15:32:57+00:00

The main shortcomings existing automation systems for water supply and sewerage facilities are analyzed. One of the most promising results increasing the efficiency technological processes using conceptual analysis (the concept of digital twins) is determined. Taking into account the cost of creating products such solutions, conceptual schemes for their practical application within the framework of the already industrial SCADA and automated process control systems with additions resource knowledge bases, laboratory information-modeling mechanisms and virtual physical and mathematical ones are substantiated and proposed.

The intelligent system for assessing postural and kinetic tremor in patients diagnosed Parkinson's disease.

2025-01-14T15:42:12+00:00

The article discusses the application of machine learning methods for the diagnosis of Parkinson's disease. Machine learning methods suitable for solving the classification problem are identified. Software is developed to perform movement detection for tremor assessment. Neural network models, logistic regression, XGBoost classifier, random forests, and support vectors were used to classify videos by the name of the exercise. The best machine learning results were obtained using the random forest classifier.

Analysis of subject areas and software products that use the finite element method

2025-01-14T15:50:16+00:00

The paper analyzes the application of mathematical calculations using the finite element method. The programs including the specified method are considered and compared. The prospects for the development of the finite element method are determined. The direction of further research is to use the research results to build new useful modifications of the finite element method, increasing its effectiveness and expanding the scope of practical application.

Specifics of defining relationships in the thesaurus in the sphere of programming.

2025-01-14T16:10:53+00:00

The article analyzes the existing models of information retrieval. The features of the thesaurus formation in the field of programming are considered. The necessity of using such a thesaurus in information search is substantiated. The structure of an information retrieval system using a thesaurus and the features of the relationship between concepts in the thesaurus ("below-above", "part-whole", associations) are described, allowing it to be attributed to formal ontologies.