Recurrent Neural Network Training With Convex Loss and Regularization Functions by Extended Kalman Filtering

• 成果类型: Article
• 署名作者: Bemporad, Alberto
• 署名单位: IMT School for Advanced Studies Lucca
• 刊物名称: IEEE TRANSACTIONS ON AUTOMATIC CONTROL
• ISSN/ISSBN: 0018-9286
• DOI: 10.1109/TAC.2022.3222750
• 发表日期: 2023
• 页码: 5661-5668
• 关键词: Extended Kalman filtering; nonlinear model predictive control; nonlinear system identification; recurrent neural networks (RNNs)

摘要

This article investigates the use of extended Kalman filtering to train recurrent neural networks with rather general convex loss functions and regularization terms on the network parameters, including $\ell _{1}$-regularization. We show that the learning method is competitive with respect to stochastic gradient descent in a nonlinear system identification benchmark and in training a linear system with binary outputs. We also explore the use of the algorithm in data-driven nonlinear model predictive control and its relation with disturbance models for offset-free closed-loop tracking.