EUVIMEDEUVIMED entdecken
Crossref

Interpretable Machine Learning: Theory, Methods, and Applications

Jiangshan Zhu

Annals of Social Sciences & Management Studies · 09.06.2026 · 13 (3)

DOI: 10.19080/asm.2026.13.555864 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Annals of Social Sciences & Management Studies, online erschienen am 09.06.2026. Verfasst von Jiangshan Zhu.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the development of artificial intelligence, its inherent “black-box” nature has sparked major concerns about accountability, fairness, and trust, which are particularly pronounced in critical domains such as healthcare, finance, and criminal justice. This paper offers a systematic and in-depth review of explainable artificial intelligence (XAI), emphasizing its key role in enhancing the interpretability and transparency of complex AI systems across various application scenarios. First, we seek to define interpretability in the context of machine learning and propose the PDR framework—Predictive, Descriptive, and Relevant—to address these issues. The PDR framework specifies three overarching evaluation objectives: predictive accuracy, descriptive accuracy, and relevance, where “relevance” is assessed with respect to human audiences. Next, we classify methods according to the models to which they apply: one category comprises interpretable methods based on self-explanatory models, and the other comprises interpretable methods based on external co-explanations. The latter is further subdivided into several branches: instance-based, SHAP-based, knowledge-graph-based, deep-learning-based, and clustering-model-based approaches. This taxonomy helps us better understand the characteristics of the models on which different explainability methods depend, thereby making it easier for researchers to select appropriate models to address interpretability problems. Finally, we review mainstream XAI applications across disciplinary and cross-disciplinary fields and discuss future challenges and developmental trends in interpretable machine learning, with the aim of promoting sustained progress in this area.

Zitieren

APA

Zhu, J. (2026). Interpretable Machine Learning: Theory, Methods, and Applications. Annals of Social Sciences & Management Studies, 13(3). https://doi.org/10.19080/asm.2026.13.555864

Vancouver

Zhu J. Interpretable Machine Learning: Theory, Methods, and Applications. Annals of Social Sciences & Management Studies. 2026;13(3). doi:10.19080/asm.2026.13.555864

BibTeX
@article{Zhu2026,
  author = {Zhu, Jiangshan},
  title = {Interpretable Machine Learning: Theory, Methods, and Applications},
  journal = {Annals of Social Sciences & Management Studies},
  year = {2026},
  volume = {13},
  number = {3},
  doi = {10.19080/asm.2026.13.555864}
}
Kostenlos starten →