Learning a Formula of Interpretability to Learn Interpretable Formulas
Authors:
Preprint
Abstract
Many risk-sensitive applications require Machine Learning (ML) models to be interpretable. Attempts to obtain interpretable models typically rely on tuning, by trial-and-error, hyper-parameters of model complexity that are only loosely related to interpretability. We show that it is instead possible to take a meta-learning approach: an ML model of non-trivial Proxies of Human Interpretability (PHIs) can be learned from human feedback, then this model can be incorporated within an ML training process to directly optimize for interpretability. We show this for evolutionary symbolic regression. We first design and distribute a survey finalized at finding a link between features of mathematical formulas and two established PHIs, simulatability and decomposability. Next, we use the resulting dataset to learn an ML model of interpretability. Lastly, we query this model to estimate the interpretability of evolving solutions within bi-objective genetic programming. We perform experiments on five synthetic and eight real-world symbolic regression problems, comparing to the traditional use of solution size minimization. The results show that the use of our model leads to formulas that are, for a same level of accuracy-interpretability trade-off, either significantly more or equally accurate. Moreover, the formulas are also arguably more interpretable. Given the very positive results, we believe that our approach represents an important stepping stone for the design of next-generation interpretable (evolutionary) ML algorithms.Keywords
All Related Versions
This publication has 30 references indexed in Scilit:
- Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AIInformation Fusion, 2020
- Time Control or Size Control? Reducing Complexity and Improving Accuracy of Genetic Programming Models2020
- Interpretable policies for reinforcement learning by genetic programmingEngineering Applications of Artificial Intelligence, 2018
- Divergence of character and premature convergence: A survey of methodologies for promoting diversity in evolutionary optimizationInformation Sciences, 2016
- Multiobjective genetic programming for maximizing ROC performanceNeurocomputing, 2014
- An interpretable classification rule mining algorithmInformation Sciences, 2013
- Better GP benchmarks: community survey results and proposalsGenetic Programming and Evolvable Machines, 2012
- Operator equalisation for bloat free genetic programming and a survey of bloat control methodsGenetic Programming and Evolvable Machines, 2011
- A multi-objective genetic programming approach to developing Pareto optimal decision treesDecision Support Systems, 2007
- Pareto-Front Exploitation in Symbolic Regression2006