Hyper-Parameter Auto-Tuning for Sparse Bayesian Learning

TL;DR

This paper introduces a neural network-based auto-tuner for hyper-parameters in sparse Bayesian learning, significantly improving convergence and recovery performance over manual tuning methods.

Contribution

It proposes a novel neural network-based auto-tuning approach for hyper-parameters in SBL, replacing empirical methods.

Findings

Improved convergence rate in hyper-parameter tuning.

Enhanced recovery performance in sparse Bayesian learning.

Neural network auto-tuner outperforms empirical auto-tuners.

Abstract

Choosing the values of hyper-parameters in sparse Bayesian learning (SBL) can significantly impact performance. However, the hyper-parameters are normally tuned manually, which is often a difficult task. Most recently, effective automatic hyper-parameter tuning was achieved by using an empirical auto-tuner. In this work, we address the issue of hyper-parameter auto-tuning using neural network (NN)-based learning. Inspired by the empirical auto-tuner, we design and learn a NN-based auto-tuner, and show that considerable improvement in convergence rate and recovery performance can be achieved.