Data-Driven Impulse Response Regularization via Deep Learning

TL;DR

This paper introduces a novel deep learning-based data-driven model for impulse response estimation in stable linear SISO systems, outperforming traditional non-parametric models by capturing more hidden patterns in input-output data.

Contribution

It presents a new flexible deep learning model for impulse response estimation, advancing beyond existing non-parametric approaches.

Findings

The deep learning model exploits more hidden patterns in data.

Experiments show improved performance over classical non-parametric models.

The approach is applicable to stable linear SISO systems.

Abstract

We consider the problem of impulse response estimation of stable linear single-input single-output systems. It is a well-studied problem where flexible non-parametric models recently offered a leap in performance compared to the classical finite-dimensional model structures. Inspired by this development and the success of deep learning we propose a new flexible data-driven model. Our experiments indicate that the new model is capable of exploiting even more of the hidden patterns that are present in the input-output data as compared to the non-parametric models.