How to address monotonicity for model risk management?
This work addresses model risk management for stakeholders requiring transparent and fair AI systems, but it is incremental as it builds on existing monotonicity concepts with a novel hybrid method.
The paper tackled the problem of ensuring accountability and fairness in transparent machine learning models by addressing monotonicity violations, proposing monotonic groves of neural additive models to achieve this while maintaining transparency, and demonstrated through empirical examples that these models are transparent, accountable, and fair.
In this paper, we study the problem of establishing the accountability and fairness of transparent machine learning models through monotonicity. Although there have been numerous studies on individual monotonicity, pairwise monotonicity is often overlooked in the existing literature. This paper studies transparent neural networks in the presence of three types of monotonicity: individual monotonicity, weak pairwise monotonicity, and strong pairwise monotonicity. As a means of achieving monotonicity while maintaining transparency, we propose the monotonic groves of neural additive models. As a result of empirical examples, we demonstrate that monotonicity is often violated in practice and that monotonic groves of neural additive models are transparent, accountable, and fair.