Abstract
A new neuron model with a tunable activation function, denoted by the TAF model, and its application are addressed. The activation function as well as the connection weights of the neuron model can be adjusted in the training process. The two-spiral problem was used as an example to show how to deduce the adjustable activation function required, and how to construct and train the network by the use of thea priori knowledge of the problem. Due to the incorporation of constraints knowna priori into the activation function, many novel aspects are revealed, such as small network size, fast learning and good performances. It is believed that the introduction of the new neuron model will pave a new way in ANN studies.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Baum, E.B., Lang, K.L., Constructing hidden units using examples and queries, inAdvances in Neural Information Processing Systems (eds. Lippman, P., Moody, J.E., Touretzky, D.S.), Vol. 3. San Mateo, CA: Morgan Kaufmann, 1991 904–910.
Chen, Q.C., Thomas, D.W., Nixon, M.S. et al., Generating-shrinking algorithm for learning arbitrary classification,Neural Networks, 1994, 7: 1477.
Fahlman, S. E., Leibiere, C., The cascade-correlation learning architecture, inAdvances in Neural Information Processing Systems (ed. Touretzky, D. S.), Vol. 2, San Mateo, CA: Morgan Kaufmann, 1990, 524–532.
Stork, D.G., Allen, J.D., How to solve the N-bit parity problem with two hidden units,Neural Networks, 1992, 5: 923.
Stork, D.G., N-bit parity problems: a reply to Brown and Korn.Neural Networks, 1993, 6: 607.
Stork, D.G., Allen, J.D., How to solve the N-bit encoder with only one hidden unit,Neurocomputing, 1993, 5: 141.
Wu, Y.S., A new approach to design a simplest ANN performing certain specific problems, inProc. ICONIP’95 Beijing, Vol. 1, Beijing: Publishing House of Electronic Industry, 1995, 477–480.
Wu, Y.S., How to choose appropriate transfer function in designing a simplest ANN for solving specific problems,Science in China, Ser. E, 1996, 39(4): 369.
Author information
Authors and Affiliations
Additional information
Project supported by the National Natural Science Foundation of China.
Rights and permissions
About this article
Cite this article
Wu, Y., Zhao, M. & Ding, X. A new kind of neuron model with a tunable activation function and its applications. Sci. China Ser. E-Technol. Sci. 40, 105–112 (1997). https://doi.org/10.1007/BF02916596
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02916596