Comparative analysis of various artificial neural network learning algorithms
Автор: Kovzhun D.A., Fedorov A.A.
Журнал: Международный журнал гуманитарных и естественных наук @intjournal
Рубрика: Технические науки
Статья в выпуске: 4-3 (91), 2024 года.
Бесплатный доступ
An artificial neural network is a methodology for developing an intelligent system that can imitate the functioning of the human brain. New algorithms are increasingly being explored to improve efficiency and accuracy. Thus, this article examines the effectiveness and feasibility of using various algorithms for training an artificial neural network in the backpropagation architecture. The learning algorithms are developed and implemented in the MATLAB environment. Various learning algorithms are analyzed to determine the accuracy of results. The analysis is performed using a house price forecasting dataset from Kaggle.
Artificial neural networks, training algorithm
Короткий адрес: https://sciup.org/170204911
IDR: 170204911 | DOI: 10.24412/2500-1000-2024-4-3-11-14