- author = "Sara Silva and Vijay Ingalalli and Susana Vinga and Joao M. B. Carreiras and Joana B. Melo and Mauro Castelli and Leonardo Vanneschi and Ivo Goncalves and Jose Caldas",
- title = "Prediction of Forest Aboveground Biomass: An Exercise on Avoiding Overfitting",
- booktitle = "Applications of Evolutionary Computing, EvoApplications 2013: EvoCOMNET, EvoCOMPLEX, EvoENERGY, EvoFIN, EvoGAMES, EvoIASP, EvoINDUSTRY, EvoNUM, EvoPAR, EvoRISK, EvoROBOT, EvoSTOC",
- year = "2013",
- month = "3-5 " # apr,
- editor = "Anna I. Esparcia-Alcazar and Antonio Della Cioppa and Ivanoe {De Falco} and Ernesto Tarantino and Carlos Cotta and Robert Schaefer and Konrad Diwold and Kyrre Glette and Andrea Tettamanzi and Alexandros Agapitos and Paolo Burrelli and J. J. Merelo and Stefano Cagnoni and Mengjie Zhang and Neil Urquhart and Kevin Sim and Aniko Ekart and Francisco {Fernandez de Vega} and Sara Silva and Evert Haasdijk and Gusz Eiben and Anabela Simoes and Philipp Rohlfshagen",
- series = "LNCS",
- volume = "7835",
- publisher = "Springer Verlag",
- address = "Vienna",
- publisher_address = "Berlin",
- pages = "407--417",
- organisation = "EvoStar",
- keywords = "genetic algorithms, genetic programming",
- isbn13 = "978-3-642-37191-2",
-
DOI = "
doi:10.1007/978-3-642-37192-9_41",
- size = "11 pages",
- abstract = "Mapping and understanding the spatial distribution of forest above ground 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 generalises 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.",
-
notes = "
EvoApplications2013 held in conjunction with EuroGP2013, EvoCOP2013, EvoBio'2013 and EvoMusArt2013",
