Neural Network Based Cellular Automata Model for Dynamic Spatial Modeling in GIS
The emphasis on calibration method of neural network (NN) based cellular automata (CA) models has been limited to back propagation (BP) mostly and not much work has been done to study the effect of different NN training methods. In this article the dynamic annealing (DA) method for training NN has been compared with BP. Also the effect of various neighborhood sizes for CA has been analyzed in the context of dynamic spatial modeling for urban growth. The model has been implemented and verified for Thane city, Maharashtra state, India as this city has higher rate of urbanization compared to other cities in the state.
KeywordsCellular automata dynamic spatial modeling neural network urban growth simulated annealing geographic information system
Unable to display preview. Download preview PDF.
- 1.Fu, S.C.: Modelling Epidemic Spread using Cellular Automata, unpublished report at Department of Computer Science and Software Engineering, The University of Western Australia (2002)Google Scholar
- 3.Jacob, N., Krishnan, R., Harmsen, K., Murali Krishna, I.V.: Spatial Dynamic Modeling Through Cellular Automata for Simulating Land Use Change Dynamics. International Journal of GeoInformatics 2(2), 23–31 (2006)Google Scholar
- 5.Li, X., Yeh, A.G.O.: Urban simulation using neural networks and cellular automata for land use planning. In: Proc. ISPRS Commission IV on Geospatial Theory, Processing and Applications, part 4, Ottawa, CA, vol. XXXIV, pp. 1682–1750 (2002)Google Scholar
- 6.Marrone, P.: The Complete Guide All you need to know about JOONE (2007), http://www.joone.org
- 9.Singh, A.K.: Modelling Land Use Land Cover Changes Using Cellular Automata, In A GIS Environment, unpublished Masters thesis at International Institute for Geo-Information Science and Earth Observation (2003)Google Scholar
- 10.Soltani, A.: Towards Modeling Urban Growth with Using Cellular Automata (CA) and GIS, University of South Australia (2005)Google Scholar