Polyhedron Attention Module: Learning Adaptive-order Interactions

Learning feature interactions can be the key for multivariate predictive modeling. ReLU-activated neural networks create piecewise linear prediction models. Other nonlinear activation functions lead to models with only high-order feature interactions, thus lacking of interpretability. Recent methods incorporate candidate polynomial terms of fixed orders into deep learning, which is subject to the issue of combinatorial explosion, or learn the orders that are difficult to adapt to different regions of the feature space. We propose a Polyhedron Attention Module (PAM) to create piecewise polynomial models where the input space is split into poly-hedrons which define the different pieces and on each piece the hyperplanes that define the polyhedron boundary multiply to form the interactive terms, resulting in interactions of adaptive order to each piece. PAM is interpretable to identify important interactions in predicting a target. Theoretic analysis shows that PAM has stronger expression capability than ReLU-activated networks. Extensive experimental results demonstrate the superior classification performance of PAM on massive datasets of the click-through rate prediction and PAM can learn meaningful interaction effects in a medical problem.