TY - GEN
T1 - Prediction of forest aboveground biomass
T2 - 16th European Conference on Applications of Evolutionary Computation, EvoApplications 2013
AU - Silva, Sara
AU - Ingalalli, Vijay
AU - Vinga, Susana
AU - Carreiras, João M.B.
AU - Melo, Joana B.
AU - Castelli, Mauro
AU - Vanneschi, Leonardo
AU - Gonçalves, Ivo
AU - Caldas, José
N1 - Silva, S., Ingalalli, V., Vinga, S., Carreiras, J. M. B., Melo, J. B., Castelli, M., ... Caldas, J. (2013). Prediction of forest aboveground biomass: An exercise on avoiding overfitting. In Applications of Evolutionary Computation - 16th European Conference, EvoApplications 2013, Proceedings (pp. 407-417). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7835 LNCS). https://doi.org/10.1007/978-3-642-37192-9-41
PY - 2013/4/5
Y1 - 2013/4/5
N2 - Mapping and understanding the spatial distribution of forest aboveground biomass (AGB) is an important and challenging task. This paper describes an exercise of predicting the forest AGB of Guinea-Bissau, West Africa, using synthetic aperture radar data and measurements of tree size collected in field campaigns. Several methods were attempted, from linear regression to different variants and techniques of Genetic Programming (GP), including the cutting edge geometric semantic GP approach. The results were compared between each other in terms of root mean square error and correlation between predicted and expected values of AGB. None of the methods was able to produce a model that generalizes well to unseen data or significantly outperforms the model obtained by the state-of-the-art methodology, and the latter was also not better than a simple linear model. We conclude that the AGB prediction is a difficult problem, aggravated by the small size of the available data set.
AB - Mapping and understanding the spatial distribution of forest aboveground biomass (AGB) is an important and challenging task. This paper describes an exercise of predicting the forest AGB of Guinea-Bissau, West Africa, using synthetic aperture radar data and measurements of tree size collected in field campaigns. Several methods were attempted, from linear regression to different variants and techniques of Genetic Programming (GP), including the cutting edge geometric semantic GP approach. The results were compared between each other in terms of root mean square error and correlation between predicted and expected values of AGB. None of the methods was able to produce a model that generalizes well to unseen data or significantly outperforms the model obtained by the state-of-the-art methodology, and the latter was also not better than a simple linear model. We conclude that the AGB prediction is a difficult problem, aggravated by the small size of the available data set.
UR - http://www.scopus.com/inward/record.url?scp=84875664067&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-37192-9-41
DO - 10.1007/978-3-642-37192-9-41
M3 - Conference contribution
AN - SCOPUS:84875664067
SN - 9783642371912
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 407
EP - 417
BT - Applications of Evolutionary Computation - 16th European Conference, EvoApplications 2013, Proceedings
Y2 - 3 April 2013 through 5 April 2013
ER -