Using the k-medoids clustering algorithm and the dynamic time warping distance between time series, we clustered the time course profiles of oilseed rape cover crop. The clustering revealed five typical clusters of crop cover profiles that differed in terms of rate of increase, lag phase and maximum value, but were largely independent of the type of crop (winter/spring oil seed rape) and the weed management regime. We then constructed predictive clustering trees (a generalized form of decision trees) that predict the weed cover profile (time series) from independent (input) variables that include the crop cover cluster, other crops descriptors and environmental variables. The approach was successful in identifying the interdependencies between the weed and crop type of vegetation.
F.02 Acquisition of new scientific knowledge
COBISS.SI-ID: 24218407We use ensembles of predictive clustering trees (PCTs) for classification of diatom images in the hierarchy of taxonomic ranks. The experiments with several state-of-the-art feature extraction techniques (and their combinations) were performed on the ADIAC database of diatom images. The results show that ensembles of PCTs have better predictive performance and are more efficient than SVMs. Furthermore, the proposed system outperforms the most widely used approaches for image annotation. Finally, we demonstrate how the system can be used by taxonomists to annotate new diatom images.
F.02 Acquisition of new scientific knowledge
COBISS.SI-ID: 25233703