A Simple and Intuitive Measure for Multicriteria Evaluation of Classi cation Algorithms

Recently there has been a growing interest in methods to assist the user in the selection of adequate algorithms for supervised classiication problems. Given the user-oriented nature of these methods , it makes sense to evaluate them on a user perspective. In this paper we sketch a simple and intuitive multicriteria measure for the evaluation of algorithms and rankings of algorithms. The input of application restrictions is done in a form that is appropriate for non-expert users. Besides providing a score, the measure can be visualized in a intuitive way. Therefore it enables an iterative evaluation process. The objective of this paper is to stimulate discussion on the adequacy of the measure for user-oriented evaluation of algorithms and rankings. Therefore, no analysis is performed, neither empirical nor theoretical.