Organic Carbon (OC) is a soil property that describes the amount of carbon broken down from plants or other life forms stored in the soil below. Traditionally to quantify organic carbon content, a soil sample is collected and an alyzed in a laboratory. This is a time-consuming and ex pensive process. Recently there has been interest in ﬁnd ing cost-efﬁcient solutions to predicting organic carbon us ing remotely sensed data. Australia-wide readings of or ganic carbon and associated input factors were used in the study. The input factors included climate, landform, Land sat Multi-Spectral Scanner bands, lithology and land use. Experiments have been undertaken using Artiﬁcial Neu ral Networks (ANNs), Multiple Linear Regression (MLR) and ensembles of ANNs. A dynamic ensemble combina tion model is presented. The ensemble assesses the per formance of the ensemble members on individual cases and weights each ensemble member relative to their perfor mance on similar cases. The weighting scheme is dynamic, changing for each case in the testing set. The ensemble out put is derived from the combination of weighted ensemble members. The results encourage further research into esti mating organic carbon using ensembles.
7 p. (p. 186-192).
Proceedings of the IASTED International Conference on Advances in Computer Science and Technology : January 23-25, 2006, Puerto Vallarta, Mexico, pp. 186-192