e-space
Manchester Metropolitan University's Research Repository

    Optimizing ADMM and Over-Relaxed ADMM Parameters for Linear Quadratic Problems

    Song, Jintao, Lu, Wenqi, Lei, Yunwen, Tang, Yuchao, Pan, Zhenkuan and Duan, Jinming (2024) Optimizing ADMM and Over-Relaxed ADMM Parameters for Linear Quadratic Problems. In: The Thirty-Eighth AAAI Conference on Artificial Intelligence, 20 February 2024 - 27 February 2024, Vancouver, Canada.

    [img]
    Preview
    Accepted Version
    Available under License In Copyright.

    Download (1MB) | Preview

    Abstract

    The Alternating Direction Method of Multipliers (ADMM) has gained significant attention across a broad spectrum of machine learning applications. Incorporating the over-relaxation technique shows potential for enhancing the convergence rate of ADMM. However, determining optimal algorithmic parameters, including both the associated penalty and relaxation parameters, often relies on empirical approaches tailored to specific problem domains and contextual scenarios. Incorrect parameter selection can significantly hinder ADMM's convergence rate. To address this challenge, in this paper we first propose a general approach to optimize the value of penalty parameter, followed by a novel closed-form formula to compute the optimal relaxation parameter in the context of linear quadratic problems (LQPs). We then experimentally validate our parameter selection methods through random instantiations and diverse imaging applications, encompassing diffeomorphic image registration, image deblurring, and MRI reconstruction.

    Impact and Reach

    Statistics

    Activity Overview
    6 month trend
    104Downloads
    6 month trend
    30Hits

    Additional statistics for this dataset are available via IRStats2.

    Altmetric

    Repository staff only

    Edit record Edit record