Seminar: Rule extraction from deep neural networks

Neural network classifiers are known to be able to learn very accurate models. In the recent past, researchers have even been able to train neural networks with multiple hidden layers (deep neural networks) more effectively and efficiently. However, the major downside of neural networks is that it is not trivial to understand the way how they derive their classification decisions. To solve this problem, there has been research on extracting better understandable rules from neural networks. However, most authors focus on nets with only one single hidden layer. The paper underlying this talk (https://www.ke.tu-darmstadt.de/publications/papers/DS16DeepRED.pdf) introduces a new decompositional algorithm – DeepRED – that is able to extract rules from deep neural networks. The evaluation of the proposed algorithm shows its ability to outperform a pedagogical baseline on several tasks, including the successful extraction of rules from a neural network realizing the XOR function.