presented in Section 5. Following that, Section 6 gives some empirical results and Sec- tion 7 concludes. 2. Interpretability The key problem with interpretability is that humans are very limited in the level [...] classifier and the descrip- tion of the local regions to be interpretable. As dis- cussed in Section 2, this implies that the user may choose any learner that he deems to be appropriate. Treating the learner [...] ferent hypothesis in both tasks may provide enough variation to significantly improve the results. 5.2. Clustering Local Examples Although in the strict sense detecting local patterns is a classification …

presented in Section 5. Following that, Section 6 gives some empirical results and Sec- tion 7 concludes. 2. Interpretability The key problem with interpretability is that humans are very limited in the level [...] classifier and the descrip- tion of the local regions to be interpretable. As dis- cussed in Section 2, this implies that the user may choose any learner that he deems to be appropriate. Treating the learner [...] ferent hypothesis in both tasks may provide enough variation to significantly improve the results. 5.2. Clustering Local Examples Although in the strict sense detecting local patterns is a classification …