Usage of training methods to parameterization of multilayer neural computing structures for technological processes
The analysis of the existing training methods of multilayer neural network of computing structures is carried out. Recommendations of selected methods usage by examples of multilayer approximation tasks for technology of beneficiation were given.
Подобные документы
An analysis of some of the main methods and approaches of machine translation, which is designed to speed up the pace of translation of multilingual text. Review of the main functions of artificial neural networks used in neural machine translation.
статья, добавлен 12.01.2023The role of analysis in enterprise management. An overview of the workflow methods used to solve cloud-based enterprise operations management tasks. Substantiation of the relevance of the introduction of cloud technologies based on serverless computing.
статья, добавлен 22.02.2023Kompyutinhove storage as one of the paradigms of periodic training neural networks through the use of rekurrentnoyi other indicators. The new approach, which focuses on learning fuzzy systems thatand its parameters are adapting their structure online.
статья, добавлен 20.01.2017Проблемы интеграции Интернета вещей с технологиями Cloud computing и Fog computing. Трехуровневая архитектура, принцип работы, преимущества и недостатки технологии Fog computing, ее сравнение с Cloud computing. Области применения технологии Fog Computing.
статья, добавлен 12.07.2022The methods for adaptive, knowledge-based learning, such as online incremental feature selection, spiking neural networks, adaptive data and model integration, and artificial life systems, particle swarm optimisation, ensembles of evolving systems.
книга, добавлен 09.04.2013The technical aspects of building a solution to our problem of predicting advertisement banner efficiency. Rectified Linear Units activation function. A simple neural network architecture, trustworthy model. Visualizing convolutional neural networks.
дипломная работа, добавлен 13.07.2020The concept of cloud computing. Types of clouds. Advantages and disadvantages of using the network technology. Examining the main parts of the computing system. Studying the protection problems of information security in corporate and public systems.
реферат, добавлен 20.03.2016Major machine translation methods designed to speed up the process of multilingual text translation by using a computer software. Analysis of the performance of rule-based machine translation, statistical models, hybrid and neural machine translation.
статья, добавлен 16.08.2022Possibilities and features of artificial neural networks. Areas of practical application of artificial neural networks. Voice interface distribution and visual serach. Learning neural networks. The following visualization. Direct error propagation.
презентация, добавлен 06.02.2019The principles of construction of the integrated modular avionics aircraft and its functional elements. Methods and means of control of IMA computer systems. Evaluate the reliability of IMA computing structures for different hardware configurations.
статья, добавлен 13.07.2022Intellectual information systems as hybrid neural networks. Using the topology of neural networks based on a fuzzy perceptron, ANFIS, TSK, Wang & Mendel’s network to solve diagnostic problems in medicine. Influence of the fuzzy rule on output signal.
статья, добавлен 22.12.2016The current state of application and development of cloud computing, the main advantages and disadvantages of their use in the states, enterprises and in scientific activity. Guidance documents in the field of cloud computing information security.
статья, добавлен 23.10.2020Development of the program, applying the technique of surface voxelization of 3D-models based on graphic processes. Review of existing methods voxelization 3D-scenes view of the art surface voxelization. Analysis and results of performance tests.
дипломная работа, добавлен 11.07.2016История развития нейросетей, характеристика их главных задач. Особенности архитектуры и параметры обучения. Краткое описание программы Neural Network Wizard. Изучение принципов работы программного эмулятора нейрокомпьютера. Сбор статистики по процессу.
реферат, добавлен 14.10.2013Разработка на основе технологии Neural Network программных комплексов для прогнозирования различных явлений, статистической обработки данных в медицинском учреждении. Система принятия решений при управлении пациентопотоком в медицинских учреждениях.
статья, добавлен 05.04.2020The classical methods of active identification and adaptation of automatic control systems (ACS) were used. Test of ACS.Transient process which allows to obtain parameters of the object transfer function and calculate the conventional adjustment knob.
статья, добавлен 30.07.2016Comprehensive consideration of the relevance of password use. Analysis of current password statistics, detailed study of methods for hacking accounts. Description of alternative authentication methods and additional steps of multi-factor authentication.
статья, добавлен 20.03.2024Application of mathematical methods - important condition in the analysis, work optimization of industrial structures. Use of the device of imitating modeling for research of production systems. The program table parameters for collecting statistics.
дипломная работа, добавлен 10.11.2014Analysis of information modeling processes and tasks, and options for improvement using extensive programming capabilities. Characteristics of the main methods of automating and speeding up the design process using programming and their properties.
статья, добавлен 30.08.2022Ubiquitous computing - filling the real world with computers. Challenging HCI Assumptions. Implications for device size as well as relationship to people. Defining the Interaction Experience. Multi-scale and distributed output. Merging Digital Worlds.
презентация, добавлен 28.12.2013The technique of drawing up algorithms on elliptic curves, scalar multiplication on Koblitz curves and their effective implementation in reconfigurable devices. The basic architecture and methods for computing scalar multipfication in Hessian form.
монография, добавлен 08.03.2014Show that computing e-th roots modulo n is easier than factoring n with currently known methods, given subexponential access to an oracle outputting the roots of numbers of the form xi + c. The strategy – a general outline. Polynomial construction.
курс лекций, добавлен 08.03.2014Задача прогнозирования временных рядов как одна из классических задач, эффективно решаемых с помощью нейронных сетей. Особенности работы с пакетом Neural Network Wizard (создание модели нейронной сети). Правила распознавания цифр на базе нейронной сети.
лабораторная работа, добавлен 20.02.2012- 24. Improving the efficiency of the solution of linear programming tasks applying the CUDA-technology
Methods for improving the efficiency of the solution of large dimensional linear programming tasks based on application of parallel data processing technology – CUDA. The specific features and algorithmic peculiarities of implementing of the program.
статья, добавлен 26.09.2016 Using Duraton-Overman test to study detailed location patterns of the forestry and agricultural industry in Russia. Data analysis package. The first article on geocoding. Methods for measuring localization. Distance-based localization measurement.
курсовая работа, добавлен 11.02.2017