Local irritation/corrosion testing strategies: Extending a decision support system by applying self-learning classifiers

dProcedures have been established and tested for the extension of a decisio n support system (DSS) for the prediction of the local irritation/corrosion potential of chemicals by using self-learning classifiers. The different a pproaches (decision trees, distances examinations in a multidimensional spa ce, k-nearest neighbour method) have been implemented, tested and evaluated independently. A combination of all of the established extension approache s Nas also developed and tested. Self-learning classifiers are constructed "automatically" by a computer i.e. they are not derived by a human expert, and thus they can be constructed with minimal effort. The classifiers prese nted here extend the existing DSS in a manner that increased significantly the predictive power of the extended system. However, automatically calcula ted results of self-learning classifiers are produced by a machine, and a m achine is incapable of explaining the toxicological relevance of the result s obtained. Thus, these results must be accepted, despite an inability to p rove their reliability. Only the mathematical correctness of the the method and the prediction rates for suitable test cases can lend some credibility to predictions produced by a computer calculating on a self-learning basis . This may not be adequate for regulatory hazard assessment purposes.