Towards Logical Specification of Statistical Machine Learning
arXiv:1907.10327 · doi:10.1007/978-3-030-30446-1_16
Abstract
We introduce a logical approach to formalizing statistical properties of machine learning. Specifically, we propose a formal model for statistical classification based on a Kripke model, and formalize various notions of classification performance, robustness, and fairness of classifiers by using epistemic logic. Then we show some relationships among properties of classifiers and those between classification performance and robustness, which suggests robustness-related properties that have not been formalized in the literature as far as we know. To formalize fairness properties, we define a notion of counterfactual knowledge and show techniques to formalize conditional indistinguishability by using counterfactual epistemic operators. As far as we know, this is the first work that uses logical formulas to express statistical properties of machine learning, and that provides epistemic (resp. counterfactually epistemic) views on robustness (resp. fairness) of classifiers.
arXiv commentSEFM'19 conference paper (full version with errors corrected)
References in corpus (7)
- Explaining and Harnessing Adversarial Examples
- Equality of Opportunity in Supervised Learning
- DeepXplore: Automated Whitebox Testing of Deep Learning Systems
- Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment
- Adversarial Attacks and Defences: A Survey
- On Formalizing Fairness in Prediction with Machine Learning
- A Formalization of Robustness for Deep Neural Networks