Charles Explorer logo
🇬🇧

Evaluation of fuzzy ILP classifier on data about car accidents

Publication at Faculty of Mathematics and Physics |
2010

Abstract

Our motivation is to classify web reports about car and other accidents according to its seriousness. For this purpose we have used a new classification method based on inductive logic programming and fuzzy logic.

In this paper we present a comparison of this method with other classifiers like J48 decision trees, SVM and multilayer perceptron. Our results show that on our dataset the fuzzy ILP method is fully comparable with the alternatives.