Explaining Neural Networks Semantically and Quantitatively
This addresses the need for interpretability in neural networks, particularly for applications requiring understanding of predictions, but it is incremental as it builds on existing explainable models.
The paper tackles the problem of explaining CNN predictions by distilling knowledge into an explainable additive model, with experimental results showing effectiveness.
This paper presents a method to explain the knowledge encoded in a convolutional neural network (CNN) quantitatively and semantically. The analysis of the specific rationale of each prediction made by the CNN presents a key issue of understanding neural networks, but it is also of significant practical values in certain applications. In this study, we propose to distill knowledge from the CNN into an explainable additive model, so that we can use the explainable model to provide a quantitative explanation for the CNN prediction. We analyze the typical bias-interpreting problem of the explainable model and develop prior losses to guide the learning of the explainable additive model. Experimental results have demonstrated the effectiveness of our method.