|
Image Processing
|
An Efficient Artificial Neural Network Training Method Through Induced Learning Retardation: Inhibited Brain Learning
Conclusions
The inhibited brain learning training method proved to be a very simple, effective and promising procedure for training artificial neural network. The procedure significantly increased the training speed of a large network. Furthermore, the tendency of a large artificial neural network to be over specific, hence have less generalizing capability, was avoided through reduced training time. The procedure generated a trained large network that can generalize and more accurate for satellite image classification.
Figure 3. Classified satellite image using the trained ANN. Image a was generated using
conventinal training ANN method (control b) and image b using with 225% deviation
and 15% clamped connection limits, respectively.
References
-
Atkinson, P.M., and Tatnall, A.R.L., 1997. Neural networks in remote sensing, International Journal of Remote Sensing, 18 (4), pp. 699-709.
- Kanellopoulos, I., and Wilkinson, G.G., 1997. Strategies and best practice for neural network image classification, International Journal of Remote Sensing, 18 (4), pp. 711-725.
- Karnin, E.D., 1990. A simple procedure for pruning back-propagation trained neural networks, IEEE Transactions on Neural Networks, 1, pp. 239-242.
- Kavsoglu, T., and Mather, P.M., 1999. Pruning artificial neural networks: an example using land cover classification of multi-sensor images, International Journal of Remote Sensing, 20 (14), pp. 2787-2803.
- Rumelhart, D.E., Hinton,G.E., and Williams, R.J., 1986. Learning internal representations by error propagation. In: Parallel Distributed Processing: Explorations in the Microstructure of Cognition, Volume 1: Foundations, edited by Rumelhart, D.E., McLelland, J.L., and the PDP Research Group (Cambridge, Massachusetts: MIT Press), pp. 318-362.
|
|
|
|
|
