New PDF release: Advances in Neural Network Research and Applications

By Chen-Feng Wu, Yu-Teng Chang, Chih-Yao Lo, Han-Sheng Zhuang (auth.), Zhigang Zeng, Jun Wang (eds.)

ISBN-10: 3642129897

ISBN-13: 9783642129896

This publication is part of the complaints of the 7th overseas Symposium on Neural Networks (ISNN 2010), hung on June 6-9, 2010 in Shanghai, China. ISNN 2010 bought quite a few submissions from approximately millions of authors in approximately forty international locations and areas throughout six continents . in accordance with the rigorous peer-reviews via this system committee individuals and the reviewers, 108 top of the range papers have been chosen for guides in Lecture Notes in electric Engineering (LNEE) complaints. those papers conceal all significant subject matters of the engineering designs and purposes of neural community examine. as well as the contributed papers, the ISNN 2010 technical application incorporated 4 plenary speeches by way of Andrzej Cichocki (RIKEN mind technology Institute, Japan), Chin-Teng Lin (National Chiao Tung college, Taiwan), DeLiang Wang (Ohio nation college, USA), Gary G. Yen (Oklahoma kingdom college, USA).

Show description

Read or Download Advances in Neural Network Research and Applications PDF

Best technique books

Read e-book online F/A-18E&F/A-18F Super Hornet in detail & scale PDF

A hugely informative and particular inside-&-out examine the trendy U. S. fighter-bomber. includes complete colour profiles, B&W line artwork, technical information, fifty one colour photographs, and dozens of Black & White photographs.

Additional resources for Advances in Neural Network Research and Applications

Example text

We insert the modeling data in the network to operate. The model order is 3, the input layer nodes number using the network is 2, hidden layer node number s is 5, and output layer nodes number is 1. After 15000 iterations, the network converges, and the steady network structure, The Study of Forecasting Model of Rock Burst for Acoustic Emission 17 connection weights and threshold are got. The fitted and measured values are compared in Table 1. Table 1. 61 Table 2. 62%. The rates of acoustic emission at 265s and 270s are predicted, and compared with the record values to test the correctness of the model.

With the rapid development of computer and artificial intelligence technology, the prediction models about the rate of Z. Zeng & J. ): Adv. , LNEE 67, pp. 21–27. com © Springer-Verlag Berlin Heidelberg 2010 22 H. Niu et al. the water content in crude oil based on ANN(artificial neural network) have been put forward recently[2]. The ANN is suitable for the non-linear prediction, and in theory it can approach the non-linear sequences in any precision. But the ANN is hard to decide the structure of the net in the application, and it like slow convergence of learning and liability of dropping into local minimum [3].

Xu From the time series X (i ) , i = 1,2, " , N , N − p samples can be constructed. 1 for learning and training, and then a stable network structure, connection weights and thresholds can be got. The time series forecasting model of geotechnical engineering is established based on artificial neural network. The input vectors are inputted into the network, using the stable network structure, connection weights and thresholds, so that the samples can be predicted. 3 Catastrophe Cusp Model and Yield Mechanism Cusp theory of the catastrophe theory is the most widely used in the range of mechanics [11].

Download PDF sample

Advances in Neural Network Research and Applications by Chen-Feng Wu, Yu-Teng Chang, Chih-Yao Lo, Han-Sheng Zhuang (auth.), Zhigang Zeng, Jun Wang (eds.)

by Paul

Rated 4.44 of 5 – based on 41 votes