Towards analogy-based explanations in machine learning

DOI10.1007/978-3-030-57524-3_17zbMATH Open1478.68272arXiv2005.12800OpenAlexW3082478769MaRDI QIDQ1997251FDOQ1997251

Authors: Eyke Hüllermeier 

Publication date: 1 March 2021

Abstract: Principles of analogical reasoning have recently been applied in the context of machine learning, for example to develop new methods for classification and preference learning. In this paper, we argue that, while analogical reasoning is certainly useful for constructing new learning algorithms with high predictive accuracy, is is arguably not less interesting from an interpretability and explainability point of view. More specifically, we take the view that an analogy-based approach is a viable alternative to existing approaches in the realm of explainable AI and interpretable machine learning, and that analogy-based explanations of the predictions produced by a machine learning algorithm can complement similarity-based explanations in a meaningful way. To corroborate these claims, we outline the basic idea of an analogy-based explanation and illustrate its potential usefulness by means of some examples.

Full work available at URL: https://arxiv.org/abs/2005.12800

Recommendations

• A survey on the explainability of supervised machine learning
• On the role of interpretive analogy in learning
• Argument based machine learning
• Analogical classifiers: a theoretical perspective
• Definitions, methods, and applications in interpretable machine learning

zbMATH Keywords

classificationmachine learninginterpretabilityrankingsimilarityanalogyexplainability

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) General topics in artificial intelligence (68T01)

Cited In (11)

• Case-based prediction -- a survey
• Using analogical proportions for explanations
• Analogy-based classifiers: an improved algorithm exploiting competent data pairs
• Galois theory for analogical classifiers
• A new paradigm for artificial intelligence based on group equivariant non-expansive operators
• Score-Based Explanations in Data Management and Machine Learning
• Multiple analogical proportions
• Decomposing analogy: a logic characterization
• A survey on the explainability of supervised machine learning
• Beneficial and harmful explanatory machine learning
• The explanation game: a formal framework for interpretable machine learning

This page was built for publication: Towards analogy-based explanations in machine learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1997251)