Factorized Structured Regression for Large-Scale Varying Coefficient Models
This work addresses the problem of applying complex statistical models to large-scale data, such as in recommender systems, though it appears incremental as it builds on existing regression and factorization methods.
The paper tackles the scalability limitations of structured regression models for large-scale varying coefficient models by proposing Factorized Structured Regression (FaStR), which combines structured additive regression and factorization in a neural network to achieve competitive estimation and prediction performance while scaling notably better than state-of-the-art techniques.
Recommender Systems (RS) pervade many aspects of our everyday digital life. Proposed to work at scale, state-of-the-art RS allow the modeling of thousands of interactions and facilitate highly individualized recommendations. Conceptually, many RS can be viewed as instances of statistical regression models that incorporate complex feature effects and potentially non-Gaussian outcomes. Such structured regression models, including time-aware varying coefficients models, are, however, limited in their applicability to categorical effects and inclusion of a large number of interactions. Here, we propose Factorized Structured Regression (FaStR) for scalable varying coefficient models. FaStR overcomes limitations of general regression models for large-scale data by combining structured additive regression and factorization approaches in a neural network-based model implementation. This fusion provides a scalable framework for the estimation of statistical models in previously infeasible data settings. Empirical results confirm that the estimation of varying coefficients of our approach is on par with state-of-the-art regression techniques, while scaling notably better and also being competitive with other time-aware RS in terms of prediction performance. We illustrate FaStR's performance and interpretability on a large-scale behavioral study with smartphone user data.