Explainable Artificial Intelligence: Understanding, Visualizing and Interpreting Deep Learning Models

TLDR

Deep learning models now achieve human‑level performance across many domains, yet their complex, black‑box nature limits transparency, especially in critical fields like medicine, prompting a growing demand for visualization and interpretation methods. The paper reviews recent advances in explainable AI and advocates for greater interpretability, presenting two methods for explaining deep learning predictions. The authors evaluate two explanation techniques—input‑sensitivity analysis and input‑variable decomposition—on three classification benchmarks. The review underscores the importance of interpretability and demonstrates that the two proposed explanation methods can clarify deep learning predictions on diverse classification tasks.

Abstract

With the availability of large databases and recent improvements in deep learning methodology, the performance of AI systems is reaching or even exceeding the human level on an increasing number of complex tasks. Impressive examples of this development can be found in domains such as image classification, sentiment analysis, speech understanding or strategic game playing. However, because of their nested non-linear structure, these highly successful machine learning and artificial intelligence models are usually applied in a black box manner, i.e., no information is provided about what exactly makes them arrive at their predictions. Since this lack of transparency can be a major drawback, e.g., in medical applications, the development of methods for visualizing, explaining and interpreting deep learning models has recently attracted increasing attention. This paper summarizes recent developments in this field and makes a plea for more interpretability in artificial intelligence. Furthermore, it presents two approaches to explaining predictions of deep learning models, one method which computes the sensitivity of the prediction with respect to changes in the input and one approach which meaningfully decomposes the decision in terms of the input variables. These methods are evaluated on three classification tasks.