THE MULTI-STAGE TRAINING METHOD FOR A CONVOLUTIONAL NEURAL NETWORK
Journal Title: Международный научный журнал "Интернаука" - Year 2018, Vol 1, Issue 6
Abstract
This paper describes methods and algorithms for training neural networks. Also, the organization of multi-stage training of the neural network, based on adaptive and genetic methods, which has been successfully applied to the convergent neural network to solve the problem of object classification.
Authors and Affiliations
Oxana Tarasenko-Klyаtchenkо, Viktoriia Buts
Keywords
Related Articles
- EP ID EP599396
- DOI -
- Views 60
- Downloads 0
SENSORY ANALYSIS OF MAYONNAISE CANTEENS DOMESTIC AND FOREIGN PRODUCTION
The article presents the results of the sensory analysis of mayonnaise canteens domestic and foreign production with the use of the profile method.
ОСОБЕННОСТИ СОЦИАЛИЗАЦИИ ДЕТЕЙ В МНОГОДЕТНЫХ СЕМЬЯХ
Данная статья рассматривает такую проблему современного общества, как особенности социализации детей в многодетных семьях. В статье проанализированы результаты социологического исследования, проведенного в г. Тюмени в хо...
ANALYSIS AND INDUSTRIAL-ECONOMIC EVALUATION OF THE ACTIVITY OF THE ORGANIZATION
In the context of competition in the markets for goods and services, it is essential to conduct a detailed analysis and accurate forecasting of the volumes of production of goods and services. Therefore, it is required a...
DETERMINATION OF STRESS STATE NEAR THE HOLES IN CYLINDRICAL SHELL
The method of determination of stresses in cylindrical shell with two holes on one guide was given. The shells model of Timoshenko was used. The method of finite elements was used. The strain distribution around the hole...
SOME ASPECTS OF THE PROTECTION OF NATURAL AND CULTURAL HERITAGE IN AZERBAIJAN
Highlighted main trends of the protection of natural and cultural heritage in Azerbaijan.