ECAI 2004 Conference Paper

[PDF] [full paper] [prev] [tofc] [next]

Instance-Based Prediction with Guaranteed Confidence

Eyke Hüllermeier

Instance-based learning (IBL) algorithms have proved to be successful in many applications. However, as opposed to standard statistical methods, a prediction in IBL is usually given without characterizing its confidence. In this paper, we propose an IBL method that allows for deriving set-valued predictions that cover the correct answer (label) with high probability. Our method makes use of a formal model of the heuristic inference principle suggesting that similar instances do have similar labels. The focus of this paper is on the prediction of numeric values (regression), even though the method is also useful for classification problems if a reasonable similarity measure can be defined on the set of classes.

Keywords: Machine Learning, Case-Based Reasoning

Citation: Eyke Hüllermeier: Instance-Based Prediction with Guaranteed Confidence. In R.López de Mántaras and L.Saitta (eds.): ECAI2004, Proceedings of the 16th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2004, pp.97-101.


[prev] [tofc] [next]


ECAI-2004 is organised by the European Coordinating Committee for Artificial Intelligence (ECCAI) and hosted by the Universitat Politčcnica de Valčncia on behalf of Asociación Espańola de Inteligencia Artificial (AEPIA) and Associació Catalana d'Intel-ligčncia Artificial (ACIA).