An optimization approach to identify the relationship between features and output of a multi-label classifier
Mammadov, Musa, Rubinov, Alex and Yearwood, John Leighton 2007, An optimization approach to identify the relationship between features and output of a multi-label classifier. In Pardalos, Panos M., Boginski, Vladimir L. and Vazacopoulos, Alkis (ed), Data Mining in Biomedicine, Springer, Boston, Mass., pp.141-167, doi: 10.1007/978-0-387-69319-4_9.
Title
An optimization approach to identify the relationship between features and output of a multi-label classifier
Multi-label classification is an important and difficult problem that frequently arises in text categorization. The accurate identification of drugs which are responsible for reactions that have occurred is one of the important problems of adverse drug reactions (ADR). In this chapter we consider the similarities of these two problems and analyze the usefulness of drug reaction relationships for the prediction of possible reactions that may occur. We also introduce a new method for the determination of responsibility for subsets of drug(s), out of all drugs taken by a particular patient, in reactions that have been observed. This method is applied for the evaluation of the level of correctness of suspected drugs reported in Cardiovascular type reactions in the ADRAC database. The problem of interaction of drugs is also considered.
Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO. If you believe that your rights have been infringed by this repository, please contact drosupport@deakin.edu.au.