Low-rank MMSE filters, Kronecker-product representation, and regularization: a new perspective
By: Daniel Gomes de Pinho Zanco , Leszek Szczecinski , Jacob Benesty and more
In this work, we propose a method to efficiently find the regularization parameter for low-rank MMSE filters based on a Kronecker-product representation. We show that the regularization parameter is surprisingly linked to the problem of rank selection and, thus, properly choosing it, is crucial for low-rank settings. The proposed method is validated through simulations, showing significant gains over commonly used methods.
Similar Papers
Shift Before You Learn: Enabling Low-Rank Representations in Reinforcement Learning
Machine Learning (CS)
Helps robots learn faster by skipping early steps.
A Probabilistic Basis for Low-Rank Matrix Learning
Machine Learning (Stat)
Improves computer guessing for missing data.
Improving Personalized Search with Regularized Low-Rank Parameter Updates
CV and Pattern Recognition
Teaches computers to find your specific things.