Towards Explainable Artificial Intelligence

By Wojciech Samek, K. Müller. Explainable AI, 2019

521 citations recorded by the source corpus.

In recent years, machine learning (ML) has become a key enabling technology for the sciences and industry. Especially through improvements in methodology, the availability of large databases and increased computational power, today’s ML algorithms are able to achieve excellent performance (at times even exceeding the human level) on an increasing number of complex tasks. Deep learning models are at the forefront of this development. However, due to their nested non-linear structure, these powerful models have been generally considered “black boxes”, not providing any information about what exactly makes them arrive at their predictions. Since in many applications, e.g., in the medical domain, such lack of transparency may be not acceptable, the development of methods for visualizing, explaining and interpreting deep learning models has recently attracted increasing attention. This introductory paper presents recent developments and applications in this field and makes a plea for a wider use of explainable learning algorithms in practice.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/21f214165bda8813c95cfa3e471e24aaf8d5776b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Unbox the black-box for the medical explainable AI via multi-modal and multi-centre data fusion: A mini-review, two showcases and beyond (2021, background): “Recently, model-agnostic based approaches attract great attention that rely on a simplified surrogate function to explain the predictions [3].”
Explainable Artificial Intelligence: a Systematic Review (2020, methodology): “In the latter case, the objective is to explicitly explain each inference of a model [130, 26, 56, 17];”
A comprehensive taxonomy for explainable artificial intelligence: a systematic survey of surveys on methods and concepts (2021, background): “[118]: A short introductory survey on visual explainable AI for researchers, giving an overview on important developments;”
Potential, challenges and future directions for deep learning in prognostics and health management applications (2020, background): “Furthermore, different approaches have been proposed to explain the behaviour of machine learning approaches post-hoc, these include (Samek and Müller, 2019): • Explaining with Surrogates • Explaining with local perturbations • Propagation-Based Approaches (Leveraging Structure) • Meta-explanations.”
Interpretable Machine Learning - A Brief History, State-of-the-Art and Challenges (2020, unknown): “Research-wise, the field is maturing in terms of methods surveys [75,41,120,96,1,6,23,15], further consolidation of terms and knowledge [42,22,82,97,88,17], and work about defining interpretability or evaluation of IML methods [74,73,95,49].”
A Survey on Neural Network Interpretability (2020, background): “[33] had a short subsection talking about “type of explanations” (including explaining learned representations, explaining individual predictions etc.”
The Future of Sensitivity Analysis: An essential discipline for systems modeling and policy support (2020, unknown): “These two terms, usually used interchangeably in the literature, refer to the ability of a model developer to make sense of why the model functions the way it does and to explain that to a user (Rudin, 2019; Samek and Müller, 2019; Roscher et al., 2020).”
Trustworthy Artificial Intelligence: A Review (2022, methodology): “In-modeling approaches work well with less complicated model families like decision trees [153], linear models [173], and rule-based models [71].”
Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future (2020, methodology): “k’s structure and training is one way to explore how it functions. Additionally, trained networks can be probed directly using techniques normally applied to brains, or those only available in models [81]. In either case, if we believe CNNs have been validated as a model of the visual system, any in”