Explainable Preference Learning: a Decision Tree-based Surrogate Model for Preferential Bayesian Optimization
This work addresses the need for more interpretable and scalable optimization methods in machine learning, particularly for applications like preference learning, though it is incremental in improving existing surrogate models.
The paper tackled the problem of interpretability and scalability in Preferential Bayesian Optimization by introducing a decision tree-based surrogate model, which outperformed Gaussian Process-based alternatives on spiky functions with only marginally lower performance on non-spiky functions, as shown in experiments on eight optimization functions and the Sushi dataset.
Current Preferential Bayesian Optimization methods rely on Gaussian Processes (GPs) as surrogate models. These models are hard to interpret, struggle with handling categorical data, and are computationally complex, limiting their real-world usability. In this paper, we introduce an inherently interpretable decision tree-based surrogate model capable of handling both categorical and continuous data, and scalable to large datasets. Extensive numerical experiments on eight increasingly spiky optimization functions show that our model outperforms GP-based alternatives on spiky functions and has only marginally lower performance for non-spiky functions. Moreover, we apply our model to the real-world Sushi dataset and show its ability to learn an individual's sushi preferences. Finally, we show some initial work on using historical preference data to speed up the optimization process for new unseen users.