Кафедри
Постійне посилання на розділhttps://repository.kpi.kharkov.ua/handle/KhPI-Press/35393
Переглянути
5 результатів
Результати пошуку
Документ Development of method for identification the computer system state based on the decision tree with multidimensional nodes(Запорізький національний технічний університет, 2022) Gavrylenko, Svitlana; Chelak, V. V.; Semenov, S. G.Context. The problem of identifying the state of a computer system is considered. The object of the research is the process of computer system state identification. The subject of the research is the methods of constructing solutions for computer system state identification. Objective. The purpose of the work is to develop a method for decision trees learning for computer system state identification. Method. A new method for constructing a decision tree is proposed, combining the classical model for constructing a decision tree and the density-based spatial clustering method (DBSCAN). The simulation results showed that the proposed method makes it possible to reduce the number of branches in the decision tree, which will increase the efficiency of identifying the state of the computer system. Belonging to hyperspheres is used as a criterion for decision-making, which enables to increase the identification accuracy due to the nonlinearity of the partition plane and to perform a more optimal adjustment of the classifier. The method is especially effective in the presence of initial data with high correlation coefficients, since it combines them into one or more multivariate criteria. An assessment of the accuracy and efficiency of the developed method for identifying the state of a computer system is carried out. Results. The developed method is implemented in software and researched in solving the problem of identifying the state of the functioning of a computer system. Conclusions. The carried out experiments have confirmed the efficiency of the proposed method, which makes it possible to recommend it for practical use in order to improve the accuracy of identifying the state of a computer system. Prospects for further research may consist in the development of an ensemble of decision trees.Документ Method of building the semantic network of distributed search in e-learning(Харківський національний університет радіоелектроніки, 2017) Kuchuk, N. G.; Artiukh, R. V.; Nechausov, A. S.The subject matter of the article is semantic networks of distributed search in e-learning. The goal is to synthesize a decision tree and a stratified semantic network that allows network intelligent agents in the e-learning to construct inference mechanisms according to the required attributes and specified relationships. The following results are obtained. The model of the base decision tree in e-learning is suggested. To simulate the decision tree in e-learning, the logic of predicates of the first order was used, which enabled making calculations both at the nodes of the tree and at its edges, and making decisions based on the results of calculations; applying partitioning operations to select individual fragments; specifying the solutions with further expanding the inference upper vertices; expanding the multi-level model vertically and horizontally. At the first stage of the model formalization, the graph of the base decision tree was constructed, whose nodes represent a substructure capable of performing an autonomous search subtask. The second stage is filling the base tree with semantic information and organizing its interaction with network intelligent agents. To provide the tree branches of decisions in e-learning with information, the process of stratified expansion of the base decision tree was suggested where the components of the decision node were detailed and the links among the received sub-units were established both on the horizontal and on the vertical levels. It is shown that in order to establish a set of goals and search problems on the studied structure, it suffices to determine: the graphs of goals and search problems for each node type; a set of edges that determine the dependence of the execution of search targets for the nodes that are not of the same type; a set of pointers that establish probable relationships for redistributing resources in accordance with the requirements of intelligent agents; communication mapping. The developed mathematical model of the base decision tree enabled a stratified expansion. Determining intensions and extensions allowed stratified semantic networks to be used for searching. Conclusions. The method of synthesizing a decision tree and a stratified semantic network is suggested; this method enables considering them as closely interrelated ones in the context of distributed search in e-learning. As a result, the process of searching and designing inference mechanisms can be formalized by the network intelligent agents according to the required attributes and given relationships.Документ Parallel implementation of the method of gradient boosting(Національний технічний університет "Харківський політехнічний інститут", 2018) Tolstoluzka, E.; Parshentsev, B.; Moroz, O.The issue of machine learning has been paying more attention in all areas of information technology in recent times. On the one hand, this is due to the rapid growth of requirements for future specialists, and on the other - with the very rapid development of information technology and Internet communications. One of the main tasks of e-learning is the task of classification. For this type of task, the method of machine learning called gradient boost is very well suited. Grading boosting is a family of powerful machine learning algorithms that have proven significant success in solving practical problems. These algorithms are very flexible and easily customized for the specific needs of the program, for example, they are studied in relation to different loss functions. The idea of boosting is the iterative process of sequential building of private models. Each new model learns based on information about errors made in the previous stage, and the resulting function is a linear combination of the whole ensemble of models, taking into account minimization of any penalty function. The mathematical apparatus of gradient boosting is well adapted for the solution of the classification problem. However, as the number of input data increases, the issue of reducing the construction time of the ensemble of decision trees becomes relevant. Using parallel computing systems and parallel programming technologies can produce positive results, but requires the development of new methods for constructing gradient boosting. The article reveals the main stages of the method of parallel construction of gradient boosting for solving the classification problem in e-learning. Unlike existing ones, the method allows to take into account the features of architecture and the organization of parallel processes in computing systems with shared and distributed memory. The method takes into account the possibility of evaluating the efficiency of building an ensemble of decision trees and parallel algorithms. Obtaining performance indicators for each iteration of the method helps to select the rational number of parallel processors in the computing system. This allows for a further reduction of the completion time of the gradient boosting. The simulation with the use of MPI parallel programming technology, the Python programming language for the architecture of the DM-MIMD system, confirms the reliability of the results. Here is an example of the organization of input data. Presented by Python is a program for constructing gradient boosting. The developed visualization of the obtained estimates of performance indicators allows the user to select the necessary configuration of the computing system.Публікація Алгоритм вибору підходу до управління логістичними витратами підприємства(НТУ "ХПІ", 2014) Ковшик, Валентин ІгоровичВ статті розглядається проблема вибору підходу до управління витратами логістичної діяльності промислових підприємств. На основі аналізу літературних джерел визначено основні підходи і методи обліку, аналізу та контролю витрат, що можуть використовуватися в логістиці такі, як методи повних та прямих витрат, управління витратами за видами діяльності тощо. Виявлено їхні основні характеристики та критерії їх застосування. Пропонується алгоритм вибору підходу до управління логістичними витратами підприємств та відповідне дерево прийняття рішень, побудоване із застосуванням ID3 алгоритму на основі отриманої інформації.Документ Формування інтегрованих груп в енергетичному машинобудуванні(Хмельницький національний університет, 2013) Курбатова (Татаринцева), Юлія ЛеонідівнаОсновну увагу в цій статті приділено обґрунтуванню раціонального вибору учасників інтегрованої групи для реалізації спільної комплектації замовлення в галузі нергетичного машинобудування. Вибір підрядника - це складна багатоетапна процедура, яка може бути здійснена за допомогою моделі формування інтегрованої групи, шляхом побудови дерева рішень, на підставі рефлексивних моделей оцінки взаємодії партнерів. Формування інтегрованої групи передбачає зв'язування підприємств шляхом вибору форми міжфірмових відносин. В статті розроблено матрицю систематизації параметрів для вибору форми міжфірмових відносин учасників інтегрованої групи.