Growth Evaluation of a Conifer Forest (Pinus Cooperí Blanco) using a Neural Net Backpropagation Trained with Distance Independent Competition Measures
Estimación del Crecimiento de un Bosque de Pino (Pinus Cooperí Blanco) por medio de una Red Neuronal Backpropagation Entrenada con Índices de Competencia Independientes de la Distancia
Graduated: Jesús Celis Porras
Centro de Investigación en Computación del IPN
Av. Juan de Dios Bátiz s/n Esq. Miguel Othón de Mendizábal
C.P. 07738 México D.F ]]>
jcelisp@yahoo.com.mx
Adviser: Juan Luis Díaz de León
Centro de Investigación en Computación del IPN
Av. Juan de Dios Bátiz s/n Esq. Miguel Othón de Mendizábal
C.P. 07738 México D.F
jdiaz@pollux.cic.ipn.mx
Co–adviser: J. Alberto Gallegos Infante
Instituto Tecnológico de Durango.
Felipe Pescador 1830 Ote. CP 34000, Durango, Dgo. México
jinfantel@starmedia.com
Graduated in September 4, 2006
Abstract
To make a decision about irregular forest handling practices is very difficult cause of some characteristics like age, natural life size diversity, and spatial distribution. A very important factor to fix growth forest is the competition about natural resources, so competition between trees should be considered to develop growth model. This is possible making use of parameters building with tree dimensions like diameter high, canopy extent, top high. These parameters are the distance independent competition measures. This research shows results product to use of backpropagation neural net trained with distance independent competition measures to forecast diameter and high growth. In this work we develop a growth model of a natural mixed forest of Pinus Cooperí Blanco, endemic specie of mountain region of Durango State, Mexico. This specie has been barely studied and is very important in wood exploitation production, because is used in timber wood production, and triplay fabrication.
Key Words: Pinus Cooperí Blanco, backpropagation neural net, independent distance competition measures.
Resumen
La toma de decisiones en el manejo de bosques irregulares se dificulta en gran medida por las características como: una alta complejidad por su diversidad de edad, tamaño y distribución espacial. Una forma de conceptuar el problema es visualizar el bosque como un ecosistema, donde su estudio se basa en las interrelaciones de los organismos y su medio ambiente. Un factor que determina el crecimiento de un bosque es la competencia que existe entre los individuos de la población por recursos, por lo que la competencia entre los individuos de un bosque debe ser considerada en el desarrollo del modelo de crecimiento. Esto se logra haciendo uso de parámetros basados en las dimensiones de los árboles; como son los índices de competencia independientes de la distancia. En esta investigación se muestran los resultados obtenidos de la utilización de una. de una red neuronal backpropagation, entrenada con índices de competencias independientes de la distancia en la predicción del crecimiento en diámetro y altura de un bosque de la especie de pino Pinus Cooperi Blanco, árbol poco estudiado, sin embargo de una gran importancia en su explotación por su aprovechamiento en la producción de madera y uso en la producción de hojas de triplay.
Palabras clave: Pinus Cooperí Blanco, red neuronal backpropagation, índices de competencia independientes de la distancia.
]]>DESCARGAR ARTÍCULO EN FORMATO PDF
References
1. Atkinson P.M. and A.R. Tatnall. Introduction: "Neural networks in remote sensing", Int. J. Remote Sensing Vol. 18, Issue No. 4, pp. 699–709, 1997. [ Links ]
2. Biging Gregory S., Dobbertin M. Evaluation of Competition Indices in Individual Tree Growth Models. Forest Sciencie, Vol 41, Issue No. 2, pp. 360–377. 1995. [ Links ]
3. Bormann, F. H and G. E. Likens. "Pattern and process in a forested ecosystem." Springer–Verlag, NY, 1979. [ Links ]
4. Bruce D. and L. C. Wensel. Modelling forest growth approaches, definitions, and problems. Ek, A.R., S.R. Shifley & T.E. Burk (Eds.), Proceedings of the IUFRO symposium on Forest growth modelling and prediction, Minneapolis, Minnesota, USDA, Forest Service General Technical Report NC–120; 1–8, 1987. [ Links ]
5. Gertner, G. The need to improve models for individual tree mortality, IN: Proc. Seventh Centre Hardwood Conf. USDA For. Serv., Carbondale, IL, pp. 59–61. 1989. [ Links ]
6. Gimblett, R.H. and G. L. Ball. Neural network architectures for monitoring and simulating changes in forest resources management. AI Applications vol.9, Issue No. 2, pp. 103–123. 1995. [ Links ]
7. Glover, G. R. and J. N. Hool. A basal area ratio predictor of loblolly pine plantation mortality. For. Sci. Vol. 25, Issue No. 2, pp. 275–282, 1979. [ Links ]
8. Guan, B. T. and G. Gertner. Using a parallel distributed processing system to model individual tree mortality. For. Sci. Vol. 37, Issue No. 3, pp. 871–885. 1991a. [ Links ]
9. Guan, B. T. and G. Gertner. Modeling red pine tree survival with and artificial neural network. For. Sci. Vol. 37, Issue No. 5, pp. 1429–1440. 1991b. [ Links ]
10. Guan and Gertner. Modeling time–varying individual tree survival probability with arandom optimization procedure: An artificial neural network approach. AI Application. Applications in Natural Resource Management Journal Vol. 9, Issue No. 2, pp. 39–52. 1995. [ Links ]
11. Guan, B. T., G. Gertner and P. Parysow. A framework for uncertainty assessment of mechanistic forest growth models: A neural network example, Ecol. Model. Vol. 98, Issue No. 1, pp. 47–58, 1997. [ Links ]
12. Hasenauer, H. and D. Merkl. Forest tree mortality simulation in uneven–aged stands using connectionist networks. In: Bulsari, A. B., and S Kallio (eds.). Neural Networks in Engineering Proc. Int. Conf. on Engineering Applications of Neural Networks (EANN'97), Stockholm, Sweden, 1997. [ Links ]
13. Hilt D.E and R.M., Teck. Individual –tree diameter growth model for Northern New England. Presented at the IUFRO Growth and Yield Modeling and Prediction Conference, Minneapolis, pp. 86–92. 1987. [ Links ]
14. Hornik, K.M. Stinchcombe, M. White, H. "Multilayer feedforward networks are universal approximators," Neural Networks, Vol. 2, Issue No. 5 pp. 359–366. 1989. [ Links ]
15. Kerson Huang. Statistical Mechanics, 2d ed. (Wiley: New York,), 1987. [ Links ]
16. Kolmogorov, A.N. On the Representation of Continuous Functions of Many Variables by Superposition of Continuous Functions of One Variable and Addition, Doklady Akademii Nauk SSR, Vol. 114, pp. 953–956. 1957. [ Links ]
17. Lek, S., M. Delacoste, P. Baran, I. Dimopoulos, J. Lauques and S. Aulagnier. Application of neural networks to modeling nonlinear relation–ships in ecology. Ecol. Model. Vol. 90, Issue No. 1, pp. 39–52. 1996. [ Links ]
18. Lev Davidovich Landau and E. M. Lifshitz. Statistical Physics (Course of Theoretical Physics, vol. 5) 2d ed. (Pergamon: Oxford), 1969. [ Links ]
19. Manzanilla B. H. Los sitios permanentes de investigación silvícola un sistema integrado para iniciarse en el cultivo de los ecosistemas forestales. Boletín Técnico No 116. SARH–INIFAP. México. 101p. 1993. [ Links ]
20. McRoberts, R.E., D.L. Schmoldt and H.M. Rauscher. Enhancing the Scientific process with artificial intelligence: Forest science applications. AI Applications vol. 5 num. 2, pp. 5–26. 1991. [ Links ]
21. Moller, M. F., "A scaled conjugate gradient algorithm for fast supervised learning," Neural Networks, Vol. 6, Issue No. 4, pp. 525–533, 1993. [ Links ]
22. Montgomery Douglas C. Design and Analysis of experiments. John Wiley and Sons, Inc 1991. [ Links ]
23. Riedmiller M. Braun H. A Direct Adaptive Method for Faster Backpropagation Learning the RPROP Algorithm. In IEEE International Conference on Neural Networks (San Francisco) Vol. 1, pp. 586 –591. IEEE, New York, 1993. [ Links ]
24. Ripley, B.D. Pattern recognition and neural networks. Cambridge Univ. Press. Cambridge, G.B, 1996. [ Links ]
25. Swingler, K. "Applying neural networks: A practical guide." Academic Press, San Diego, CA, pp. 304, 1996. [ Links ]
26. Valles Gándara A.G., Torres Rojo J.M., Velázquez M. A., Rodriguez F. C. Relación de Nueve índices de competencia Con el crecimiento en diámetro de Pinus Cooperí Blanco. Agrociencia, Vol. 32, Issue No. 3, pp. 255–259. julio–septiembre, 1998. [ Links ]
27. Valles Gándara A.G., Gonzáles Laredo R.F., Gallegos Infante A., Torres Rojo J.M.,Návar Chaidez J.J., Rocha Fuentes M. Nuevos índices de competencia independientes de la distancia para predecir el crecimiento en diámetro y altura del Pinus Cooperí Blanco. RECURSOS FORESTALES_ AGROFAZ Vol. 5 Num. 1, 2005. [ Links ]
28. Valentine, H. A carbon balance model of stand growth: A derivation employing pipe–model theory and the self–thinning rule. Ann. Bot–London. Vol. 62, Issue No. 4, pp. 389–396, 1988. [ Links ]
29. Van Kampen N. G. Stochastic Processes in Physics and Chemistry (North–Holland, Amsterdam), 1981. [ Links ]
30. Weisbuch Gérard. Complex Systems Dynamics. Addison–Wesley, Reading, Mass., 1991. [ Links ]
31. Whittaker, R. Communities and Ecosystems. New York: MacMillan Press, 2nd edition, 1975. [ Links ]
32. Zhang L. "Cross–validation of non–linear growth functions for modeling tree height–diameter relationships." Ann. Bot.–London Vol. 79, Issue No. 3, pp. 251–257, 1997. [ Links ] ]]>