# Information Theoretic Evaluation of Privacy-Leakage, Interpretability, and Transferability for Trustworthy AI

arXiv: https://arxiv.org/abs/2106.06046 · HTML: https://mlsanity.com/paper/2106.06046/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 42 references that are in the map.
- First posted on arXiv 6 Jun 2021.

## Builds on (most cited first, 7 of 7)

- [Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data](https://mlsanity.com/paper/1610.05755/) (2016)
- [The Ethics of AI Ethics -- An Evaluation of Guidelines](https://mlsanity.com/paper/1903.03425/) (2019)
- [Efficient Learning of Domain-invariant Image Representations](https://mlsanity.com/paper/1301.3224/) (2013)
- [Learning an Invariant Hilbert Space for Domain Adaptation](https://mlsanity.com/paper/1611.08350/) (2016)
- [Optimal Bayesian Transfer Learning](https://mlsanity.com/paper/1801.00857/) (2018)
- [Membership-Mappings for Data Representation Learning: Measure Theoretic Conceptualization](https://mlsanity.com/paper/2104.07060/) (2021)
- [Differentially Private Transferrable Deep Learning with Membership-Mappings](https://mlsanity.com/paper/2105.04615/) (2021)

## Abstract (arXiv)

In order to develop machine learning and deep learning models that take into account the guidelines and principles of trustworthy AI, a novel information theoretic trustworthy AI framework is introduced. A unified approach to "privacy-preserving interpretable and transferable learning" is considered for studying and optimizing the tradeoffs between privacy, interpretability, and transferability aspects. A variational membership-mapping Bayesian model is used for the analytical approximations of the defined information theoretic measures for privacy-leakage, interpretability, and transferability. The approach consists of approximating the information theoretic measures via maximizing a lower-bound using variational optimization. The study presents a unified information theoretic approach to study different aspects of trustworthy AI in a rigorous analytical manner. The approach is demonstrated through numerous experiments on benchmark datasets and a real-world biomedical application concerned with the detection of mental stress on individuals using heart rate variability analysis.