The Decoupled Extended Kalman Filter for Dynamic Exponential-Family Factorization Models

Motivated by the needs of online large-scale recommender systems, we specialize the decoupled extended Kalman filter (DEKF) to factorization models, including factorization machines, matrix and tensor factorization, and illustrate the effectiveness of the approach through numerical experiments on synthetic and on real-world data. Online learning of model parameters through the DEKF makes factorization models more broadly useful by (i) allowing for more flexible observations through the entire exponential family, (ii) modeling parameter drift, and (iii) producing parameter uncertainty estimates that can enable explore/exploit and other applications. We use a different parameter dynamics than the standard DEKF, allowing parameter drift while encouraging reasonable values. We also present an alternate derivation of the extended Kalman filter and DEKF that highlights the role of the Fisher information matrix in the EKF.

Automated summary

This study focuses on applying DEKF to factorization models for online large-scale recommender systems, demonstrating through numerical experiments the effectiveness of the approach in making factorization models more broadly useful. By using a different parameter dynamics and highlighting the role of Fisher information matrix in the EKF, this method allows for parameter drift while encouraging reasonable values.