Abstract
The aim of the current study is to assess the suitability of two Associative Memory (AM) models to character recognition problems. The two AM models under scrutiny are a One-Shot AM (OSAM) and an Exponential Correlation AM (ECAM). We compare these AMs on the resultant features of their architectures, including recurrence, learning and the generation of domains of attraction. We illustrate the impact of each of these features on the performance of each AM by varying the training set size, introducing noisy data and by globally transforming symbols. Our results show that each system is suited to different character recognition problems.
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
Zurada, J.: Introduction to Artificial Neural Systems. West Publishing Company, St. Paul, MN (1992)
Skapura, D.M.: Building Neural Networks. Addison Wesley Publishing Company, Reading (1995)
Serbedzija, N.B.: Simulating artificial neural networks on parallel architectures. Computer 29, 56–70 (1996)
Tan, C.L., Quah, T.S., Teh, H.H.: An artificial neural network that models human decision making. Computer 29, 45–54 (1996)
Setiono, R., Liu, H.: Symbolic representation of neural networks. Computer 29, 71–77 (1996)
Shang, Y., Wah, B.: Global optimisation for neural network training. Computer 29, 45–54 (1996)
Jain, A.K., Mao, J., Mohiuddin, K.M.: Artificial neural networks: A tutorial. Computer 29, 31–44 (1996)
Kohonen, T.: Correlation matrix memories. IEEE Trans. Comput. 21, 353–359 (1972)
Kohonen, T.: Associative Memory: A System-Theoretical Approach. Springer, New-York (1977)
Hopfield, J.J.: Neural networks and physical systems with emergent collective computational abilities. Nat. Acad. Sci. 79, 2554–2558 (1982)
Hopfield, J.J.: Neurons with graded response have collective computational properties like those of two-state neurons. Nat. Acad. Sci. 81, 3088–3092 (1984)
Wu, Y., Pados, D.: A feedforward bidirectional associative memory. IEEE Trans. Neural Networks 11, 859–866 (2000)
Wu, Y., Batalama, N.: An efficient learning algorithm for associative memories. IEEE Trans. on Neural Networks 11, 1058–1066 (2000)
Chieuh, T., Goodman, R.M.: Recurrent correlation associative memories. IEEE Trans. on Neural Networks 2, 275–284 (1991)
Chou, P.A.: The capacity of the kanerva associative memory. IEEE Trans. on Information Theory 35, 281–298 (1989)
Hamming, R.W.: Error detecting and error correcting codes. The Bell System Technical Journal 26, 147–160 (1950)
DasGupta, B., Siegelmann, H.T., Sontag, E.D.: On the complexity of training neural networks with continuous activation functions. IEEE Transactions on Neural Networks 6, 1490–1504 (1995)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
McEnery, O., Cronin, A., Kechadi, T., Geiselbrechtinger, F. (2004). Suitability of Two Associative Memory Neural Networks to Character Recognition. In: Webb, G.I., Yu, X. (eds) AI 2004: Advances in Artificial Intelligence. AI 2004. Lecture Notes in Computer Science(), vol 3339. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-30549-1_76
Download citation
DOI: https://doi.org/10.1007/978-3-540-30549-1_76
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-24059-4
Online ISBN: 978-3-540-30549-1
eBook Packages: Computer ScienceComputer Science (R0)