Abstract
The training of neural networks occurs instantaneously with Kak’s corner classification algorithm CC4. It is based on prescriptive learning, hence is extremely fast compared with iterative supervised learning algorithms such as backpropagation. This paper shows that the Kak algorithm is hardware friendly and is especially suited for implementation in reconfigurable computing using fine grained parallelism. We also demonstrate that on-line learning with the algorithm is possible through dynamic evolution of the topology of a Kak neural network.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Hush, D.R. and B.G. Horne, Progress in supervised neural networks: what’ s new since Lippmannn? IEEE Signal Processing Magazine, IEEE Press, 1993. 10(1), p. 8–39.
Eldredge, J.G. and B.L. Hutchings, Density Enhancement of a Neural Networks Using FPGAs and Run-Time Reconfiguration. Proc. of IEEE Workshop on FPGAs for Custom Computing Machines, IEEE Computer Society Press, 1994, p. 180–188.
Hadley, J.D. and B.L. Hutchings, Design Methodologies for Partially Reconfigured Systems. Proc. of IEEE Symposium on FPGAs for Custom Computing Machines, IEEE Computer Society Press, 1995, p. 78–84.
Lysaght, P., et al., Artificial Neural Network Implementation on a Fine-Grained FPGA. Proc. of the 4th International Workshop on Field-Programmable Logic and Applications, LNCS 849, Springer, 1994, p. 421–431.
Gschwind, M., et al., Space Efficient Neural Network Implementation. Proc. of the 2nd ACM Symposium on Field-Programmable Gate Arrays, ACM, 1994, p. 23–28.
Kak, S.C., On generalization by neural networks. Journal of Information Sciences, Elsevier Science Inc., 1998. 111, p. 293–302.
Tong, K.-W. and S.C. Kak, A New Corner Classification Approach to Neural Network Training. Journal of Circuits, Systems, Signal Processing, Burkh auser Boston, 1998. 17, p. 459–469.
Raina, P., Comparison of learning and generalization capabilities of the Kak and the backpropagation algorithms. Journal of Information Sciences, Elsevier Science Inc., 1994. 81, p. 261–274.
Milne, G., et al., Realising massively concurrent systems on the SPACE Machines. Proc. of IEEE Workshop on FPGAs for Custom Computing Machines, IEEE Computer Society Press, 1993, p. 26–32.
Gunther, B.K., SPACE 2 as a Reconfigurable Stream Processor. Proc. of the 4th Australian Conference on Parallel and Real-Time Systems, Springer, 1997, p. 74–84.
Xilinx, XC6000 FPGAs, 1997, Xilinx, http://www.xilinx.com.
Xilinx, Virtex-E 1.8V Extended Memory FPGAs, 2000, Xilinx, http://www.xilinx.com.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2000 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Zhu, J., Milne, G. (2000). Implementing Kak Neural Networks on a Reconfigurable Computing Platform. In: Hartenstein, R.W., Grünbacher, H. (eds) Field-Programmable Logic and Applications: The Roadmap to Reconfigurable Computing. FPL 2000. Lecture Notes in Computer Science, vol 1896. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-44614-1_29
Download citation
DOI: https://doi.org/10.1007/3-540-44614-1_29
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-67899-1
Online ISBN: 978-3-540-44614-9
eBook Packages: Springer Book Archive