About the appropriate neural network size for the engineering applications

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21220%2F23%3A00368939" target="_blank" >RIV/68407700:21220/23:00368939 - isvavai.cz</a>
Result on the web
<a href="https://www.kme.zcu.cz/compmech/download/proceedings/CM2023_Conference_Proceedings.pdf" target="_blank" >https://www.kme.zcu.cz/compmech/download/proceedings/CM2023_Conference_Proceedings.pdf</a>
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
About the appropriate neural network size for the engineering applications
Original language description
Deep learning approaches became very popular in recent years. In terms of computational effectivity and time required for the learning process, number of degrees of freedom in the proposed neural network plays singificant role. Thus, an apriori information about the appropriate neural network size for a given task could be very promising tool in machine learning tasks. In the contrast to the standard machine learning approaches aimed to deep learning, present contribution deals with shallow higher order networks. Basics of networks are introduced and comparisons between different neural network architectures to capture more demanding engineering task is presented. Basic idea of the neural network size apriori estimation for the task is discussed.
Czech name
—
Czech description
—

Project
<a href="/en/project/EF16_019%2F0000826" target="_blank" >EF16_019/0000826: Center of Advanced Aerospace Technology</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Similar results(10)