In this paper, we tackle the problem of multi-rules based classification for evidential data, i.e., data where imperfection is modeled through the Evidence theory. In this setting, a new algorithm called EviRC is introduced. This method uses different pruning techniques to omit irrelevant rules and defines a new matching criteria between the rules and the instance to classify. The selected rules are then combined using the powerful combination rules of the Evidence theory. Extensive experiments were conducted on several data sets in order to evaluate the proposed method. The experiments produce interesting results in term of classification quality.
