iNNvestigate neural networks!
Maximilian Albera b, Sebastian Lapuschkina c d, Philipp Seegerera b, Miriam Hägelea b, Kristof T. Schütta b, Grégoire Montavona b, Wojciech Sameka d, Klaus‐Robert Müllere, Sven Dähneb f, Pieter-Jan Kindermansa b f
- aTechnische Universität Berlin , Machine Learning Group 10623 Berlin , Germany
- bTechnical University of Berlin
- cFraunhofer Heinrich Hertz Institute , Video Coding and Analytics 10587 Berlin , Germany
- dFraunhofer Institute for Telecommunications, Heinrich Hertz Institute
- eDepartment of Brain and Cognitive Engineering
- fMax Planck Institute for Informatics 66123 Saarbrücken , Germany
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
In recent years, deep neural networks have revolutionized many application domains of machine learning and are key components of many critical decision or predictive processes. Therefore, it is crucial that domain specialists can understand and analyze actions and pre- dictions, even of the most complex neural network architectures. Despite these arguments neural networks are often treated as black boxes. In the attempt to alleviate this short- coming many analysis methods were proposed, yet the lack of reference implementations often makes a systematic comparison between the methods a major effort. The presented library iNNvestigate addresses this by providing a common interface and out-of-the- box implementation for many analysis methods, including the reference implementation for PatternNet and PatternAttribution as well as for LRP-methods. To demonstrate the versatility of iNNvestigate, we provide an analysis of image classifications for variety of state-of-the-art neural network architectures.