The experiments demonstrate that FRANN compares favourably with FLORA4 in the presence of concept drift. Learning is possible from examples described by symbolic as well as by numeric attributes, and because of its representation formalism (RBF networks, which realize a kind of prototype weighting scheme) FRANN is particularly effective in capturing concepts with nonlinear boundaries.
CITATION STYLE
Kubat, M., & Widmer, G. (1995). Adapting to drift in continuous domains (Extended abstract). In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 912, pp. 307–310). Springer Verlag. https://doi.org/10.1007/3-540-59286-5_74
Mendeley helps you to discover research relevant for your work.