Protocol to explain support vector machine predictions via exact Shapley value computation

• Published in: STAR protocols Vol. 5; no. 2; p. 103010
• Main Authors: Mastropietro, Andrea, Bajorath, Jürgen
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 21.06.2024; Elsevier
• Subjects: Algorithms; Bioinformatics; Chemistry; Computer sciences; Game Theory; Humans; Machine Learning; Protocol; Software; Support Vector Machine; Bioinformatics; Chemistry; Computer sciences
• ISSN: 2666-1667; 2666-1667
• DOI: 10.1016/j.xpro.2024.103010

Shapley values from cooperative game theory are adapted for explaining machine learning predictions. For large feature sets used in machine learning, Shapley values are approximated. We present a protocol for two techniques for explaining support vector machine predictions with exact Shapley value computation. We detail the application of these algorithms and provide ready-to-use Python scripts and custom code. The final output of the protocol includes quantitative feature analysis and mapping of important features for visualization. For complete details on the use and execution of this protocol, please refer to Feldmann and Bajorath1 and Mastropietro et al.2 [Display omitted] •Guidance on calculating exact Shapley values for SVMs using the Tanimoto kernel•Instructions for calculating exact Shapley values for SVMs using RBF kernels•Steps for using Shapley values to quantify feature importance and explain predictions•Details provided on the practical use of SVETA and SVERAD Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics. Shapley values from cooperative game theory are adapted for explaining machine learning predictions. For large feature sets used in machine learning, Shapley values are approximated. We present a protocol for two techniques for explaining support vector machine predictions with exact Shapley value computation. We detail the application of these algorithms and provide ready-to-use Python scripts and custom code. The final output of the protocol includes quantitative feature analysis and mapping of important features for visualization.