Adaptive, evolving, hybrid connectionist systems for image pattern recognition
Kasabov, Nikola; Israel, Steven; Woodford, Brendon J

View/ Open
Cite this item:
Kasabov, N., Israel, S., & Woodford, B. J. (1999). Adaptive, evolving, hybrid connectionist systems for image pattern recognition (Information Science Discussion Papers Series No. 99/08). University of Otago. Retrieved from http://hdl.handle.net/10523/964
Permanent link to OUR Archive version:
http://hdl.handle.net/10523/964
Abstract:
The chapter presents a new methodology for building adaptive, incremental learning systems for image pattern classification. The systems are based on dynamically evolving fuzzy neural networks that are neural architectures to realise connectionist learning, fuzzy logic inference, and case-based reasoning. The methodology and the architecture are applied on two sets of real data—one of satellite image data, and the other of fruit image data. The proposed method and architecture encourage fast learning, life-long learning and on-line learning when the system operates in a changing environment of image data.
Date:
1999-05
Publisher:
University of Otago
Pages:
21
Series number:
99/08
Keywords:
image classification; case-based reasoning; Evolving Fuzzy Neural Networks
Research Type:
Discussion Paper
Notes:
Please note that this is a searchable PDF derived via optical character recognition (OCR) from the original source document. As the OCR process is never 100% perfect, there may be some discrepancies between the document image and the underlying text. Searching and selecting the text of this PDF may also not work in all viewers; for example, they have been found to not work in Apple's Preview application. We therefore recommend Adobe Reader for viewing and searching this PDF.
Collections
- Knowledge Engineering Laboratory [25]
- Surveying [117]
- Information Science [486]
- Discussion Paper [438]