A T-S Fuzzy neural network (FNN) is a combination of a neural network and a T-S fuzzy system. It not only can mimic the human brain logic thinking, but also has the ability of artificial neural networks, thus it is used widely in recognition. The music signal data is used in simulation. To gain a better recognition result, the error recognition data of traditional T-S FNN is analyzed by Cluster analysis. The analysis results show that this kind of data has higher ambiguity, and would generate great effect on the correct recognition rate. Aiming at this kind of error data, an improved T-S fuzzy neural network is proposed in this paper. The improvement is mainly on the learning rate and the parameter of membership function. This improved algorithm obtains a higher accurate recognition rate for the error recognition data. Compared with the traditional algorithm, the recognition rate is increased by about 10%.
[1]
R. Lippmann,et al.
An introduction to computing with neural nets
,
1987,
IEEE ASSP Magazine.
[2]
Yoshiki Uchikawa,et al.
On fuzzy modeling using fuzzy neural networks with the back-propagation algorithm
,
1992,
IEEE Trans. Neural Networks.
[3]
Michio Sugeno,et al.
Fuzzy identification of systems and its applications to modeling and control
,
1985,
IEEE Transactions on Systems, Man, and Cybernetics.
[4]
D. Signorini,et al.
Neural networks
,
1995,
The Lancet.
[5]
Chuen-Tsai Sun,et al.
Neuro-fuzzy modeling and control
,
1995,
Proc. IEEE.