Skip to Main Content
This paper presents an empirical evaluation of common vector based methods and some extensions in a particular and difficult domain corresponding to the characterization of pharmacological properties from their chemical structure for automatic drug classification problems. Several classic pattern classification methods have already been applied to this problem with promising results. In particular, it has been shown that selection of appropriate variables plays a crucial role. In this work, classification methods that explicitly look for appropriate and reduced representation spaces are considered in this particular context. Comparative experiments considering other state-of-the-art approaches in this domain are carried out.