Abstract
Neural networks are important modeling tools to implement intelligent behaviour in a wide variety of phenomena. We introduce the concept of concurrent synchrony in spikes to enable the efficient representation of neural networks to process sensory stimuli. Using different sensory modalities, we show that information processing from stimuli can be represented compactly. This approach aims at introducing homeostasis into the behavior of neural populations in order to construct diverse and sophisticated control rules without increasing network complexity.
Original language | English |
---|---|
Pages (from-to) | 304-311 |
Number of pages | 8 |
Journal | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Volume | 8834 |
Publication status | Published - 2014 Jan 1 |
Externally published | Yes |
Keywords
- Concurrency
- Neural representation
- Spiking networks
- Synchrony
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)