p.575
p.579
p.584
p.589
p.594
p.599
p.603
p.609
p.615
Studies of an Improved LMBP Model for Predicting Potential Distribution of Insects
Abstract:
An improved LMBP algorithm model was presented based on analysing the shortcomings of back propagation (BP)neural network and discussing the idea of Levenberg-Marquardt (LM)algorithm. The simulating data proved that the improved LMBP model was able to overcome its local minimum and increase the converging speed in comparison with BP algorithm. Meanwhile this model was applied to predicting the potential distribution of insects. Here we take an example of oriental fruit fly Bactrocera dorsalis. The experimental results show that the algorithm model has a capacity of learning and can solve the forecasting problem in potential distribution of oriental fruit fly. The algorithm model is better than other traditional method and is very useful and efficient in practice
Info:
Periodical:
Pages:
594-598
Citation:
Online since:
January 2012
Authors:
Keywords:
Price:
Сopyright:
© 2012 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: