Scalable Higher-Order Tensor Product Spline Models
This addresses the problem of limited model complexity in large-scale, interpretable machine learning for practitioners needing both scalability and interaction modeling.
The paper tackled the challenge of incorporating higher-order interactions in scalable, interpretable semi-parametric regression models without prohibitive computational costs, proposing a factorization method that achieves computational costs proportional to models without interactions.
In the current era of vast data and transparent machine learning, it is essential for techniques to operate at a large scale while providing a clear mathematical comprehension of the internal workings of the method. Although there already exist interpretable semi-parametric regression methods for large-scale applications that take into account non-linearity in the data, the complexity of the models is still often limited. One of the main challenges is the absence of interactions in these models, which are left out for the sake of better interpretability but also due to impractical computational costs. To overcome this limitation, we propose a new approach using a factorization method to derive a highly scalable higher-order tensor product spline model. Our method allows for the incorporation of all (higher-order) interactions of non-linear feature effects while having computational costs proportional to a model without interactions. We further develop a meaningful penalization scheme and examine the induced optimization problem. We conclude by evaluating the predictive and estimation performance of our method.