Mathematical Foundations and Optimization Strategies for Explainable AI: A Comprehensive Study

Explainable Artificial Intelligence (XAI) is important to make AI systems transparent, fair, and trustworthy. Modern deep learning models are powerful but often work like “black boxes,” making it difficult to understand how they make decisions. This paper studies the mathematical ideas that help improve explainability in AI. It focuses on important areas like linear algebra, calculus, optimization, sparse models, and information theory. We review methods such as gradient-based explanations, Shapley values, simple surrogate models, and multi-objective learning. These methods help show which features affect a model’s prediction and why. Experiments on benchmark datasets show that we can increase transparency without losing much accuracy. The results highlight that strong mathematical techniques are essential for creating reliable and ethical AI systems that people can trust.

Explainable AI, Mathematical Foundations, Optimization, Model Interpretability

1. M. T. Ribeiro, S. Singh, and C. Guestrin, "“Why Should I Trust You?”: Explaining the Predictions of Any Classifier," in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2016, pp. 1135-1144. (Note: Page numbers are often included for conference papers if known) [Google Scholar] [Crossref]

2. M. Sundararajan, A. Taly, and Q. Yan, "Axiomatic Attribution for Deep Networks," in Proceedings of the 34th International Conference on Machine Learning (ICML), 2017, pp. 3319-3328. [Google Scholar] [Crossref]

3. S. Lundberg and S.-I. Lee, "A Unified Approach to Interpreting Model Predictions," in Advances in Neural Information Processing Systems (NeurIPS), 2017, vol. 30. (SHAP) [Google Scholar] [Crossref]

4. A. Holzinger, G. Langs, H. Denk, et al., "Explainable Artificial Intelligence — Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI," in Lecture Notes in Computer Science, vol. 12278, pp. 1–18. Springer, Cham, 2020. [Google Scholar] [Crossref]

5. R. Saleem, D. Hurtado, and M. A. Ser, "Explaining Deep Neural Networks: A Survey on Global Interpretation Methods," Neurocomputing, vol. 503, pp. 67–91, 2022. [Google Scholar] [Crossref]

6. N. Balasubramaniam, J. Morio, and X. Qu, "Transparency and explainability of AI systems: From ethical principles to practical implementations," Journal of Information, Communication and Ethics in Society, 2023. [Google Scholar] [Crossref]

7. A. V. Ponce-Bobadilla, V. Schmitt, C. S. Maier, S. Mensing, and S. Stodtmann, "Practical Guide to SHAP Analysis: Explaining Supervised Machine Learning Model Predictions in Drug Development," Clinical and Translational Science, 2024. [Google Scholar] [Crossref]

8. T. Feng, Z. Zhou, J. Tarun, and V. N. Nair, "Comparing Baseline Shapley and Integrated Gradients for Local Explanation: Some Additional Insights," arXiv preprint arXiv:2208.06096, 2022. [Google Scholar] [Crossref]

9. G. Pavlidis, "Unlocking the Black Box: Analysing the EU Artificial Intelligence Act’s Framework for Explainability in AI," arXiv preprint arXiv:2502.14868, 2025. [Google Scholar] [Crossref]

10. A. Hummel, H. Burden, S. Stenberg, et al., "The EU AI Act, Stakeholder Needs, and Explainable AI: Aligning Regulatory Compliance in a Clinical Decision Support System," arXiv preprint arXiv:2505.20311, 2025. [Google Scholar] [Crossref]

11. P. Hermosilla, et al., "Explainable AI for Forensic Analysis: A Comparative Study of SHAP and LIME in Intrusion Detection Systems," Applied Sciences, vol. 15, no. 13, 2025. [Google Scholar] [Crossref]

12. N. Díaz-Rodríguez, C. Delgado-Bonal, G. Klambauer, et al., "Connecting the Dots in Trustworthy Artificial Intelligence," Information Fusion, 2023. [Google Scholar] [Crossref]

• Interplay of Students’ Emotional Intelligence and Attitude toward Mathematics on Performance in Grade 10 Algebra
• Numerical Simulation of Fitzhugh-Nagumo Dynamics Using a Finite Difference-Based Method of Lines
• Fixed Point Theorem in Controlled Metric Spaces
• Usage of Moving Average to Heart Rate, Blood Pressure and Blood Sugar
• Exploring Algebraic Topology and Homotopy Theory: Methods, Empirical Data, and Numerical Examples