Szczegóły publikacji
Opis bibliograficzny
Temporal coding of neural stimuli / Adrian HORZYK, Krzysztof GOŁDON, Janusz A. Starzyk // W: Artificial neural networks and machine learning - ICANN 2019 : workshop and special sessions : 28th International Conference on Artificial Neural Networks : Munich, Germany, September 17–19, 2019 : proceedings / eds. Igor V. Tetko, [et al.]. — Cham : Springer Nature Switzerland, cop. 2019. — (Lecture Notes in Computer Science ; ISSN 0302-9743 ; 11731). — ISBN: 978-3-030-30492-8; e-ISBN: 978-3-030-30493-5. — S. 607–621. — Bibliogr. s. 620–621, Abstr. — Publikacja dostępna online od: 2019-09-09
Autorzy (3)
- AGHHorzyk Adrian
- AGHGołdon Krzysztof
- Starzyk Janusz A.
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 124590 |
|---|---|
| Data dodania do BaDAP | 2019-09-25 |
| Tekst źródłowy | URL |
| DOI | 10.1007/978-3-030-30493-5_56 |
| Rok publikacji | 2019 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Wydawca | Springer |
| Konferencja | International Conference on Artificial Neural Networks 2019 |
| Czasopismo/seria | Lecture Notes in Computer Science |
Abstract
Contemporary artificial neural networks use various metrics to code input data and usually do not use temporal coding, unlike biological neural systems. Real neural systems operate in time and use the time to code external stimuli of various kinds to produce a uniform internal data representation that can be used for further neural computations. This paper shows how it can be done using special receptors and neurons which use the time to code external data as well as internal results of computations. If neural processes take different time, the activation time of neurons can be used to code the results of computations. Such neurons can automatically find data associated with the given inputs. In this way, we can find the most similar objects represented by the network and use them for recognition or classification tasks. Conducted research and results prove that time space, temporal coding, and temporal neurons can be used instead of data feature space, direct use of input data, and classic artificial neurons. Time and temporal coding might be an important branch for the development of future artificial neural networks inspired by biological neurons.
