Dynamic Regret Bounds for Constrained Online Nonconvex Optimization Based on Polyak-Lojasiewicz Regions

被引:3
|
作者
Mulvaney-Kemp, Julie [1 ]
Park, SangWoo [1 ]
Jin, Ming [2 ]
Lavaei, Javad [1 ]
机构
[1] Univ Calif Berkeley, Dept Ind Engn & Operat Res, Berkeley, CA 94720 USA
[2] Virginia Tech, Dept Elect & Comp Engn, Blacksburg, VA 24061 USA
来源
IEEE TRANSACTIONS ON CONTROL OF NETWORK SYSTEMS | 2023年 / 10卷 / 02期
关键词
Heuristic algorithms; Optimization; Loss measurement; Target tracking; Network systems; Convergence; Control systems; Adversarial machine learning; dynamic regret; nonconvex optimization; online optimization; optimization methods; randomized algorithms; time-varying systems; STABILITY; SYSTEMS;
D O I
10.1109/TCNS.2022.3203798
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Online optimization problems are well understood in the convex case, where algorithmic performance is typically measured relative to the best fixed decision. In this article, we shed light on online nonconvex optimization problems in which algorithms are evaluated against the optimal decision at each time using the more useful notion of dynamic regret. The focus is on loss functions that are arbitrarily nonconvex but have global solutions that are slowly time-varying. We address this problem by first analyzing the region around the global solution at each time to define time-varying target sets, which contain the global solution and exhibit desirable properties under the projected gradient descent algorithm. All points in a target set satisfy the proximal Polyak-Lojasiewicz inequality, among other conditions. Then, we introduce two algorithms and prove that the dynamic regret for each algorithm is bounded by a function of the temporal variation in the optimal decision. The first algorithm assumes that the decision maker has some prior knowledge about the initial objective function and may query the gradient repeatedly at each time. This algorithm ensures that decisions are within the target set at every time. The second algorithm makes no assumption about prior knowledge. It instead relies on random sampling and memory to find and then track the target sets over time. In this case, the landscape of the loss functions determines the likelihood that the dynamic regret will be small. Numerical experiments validate these theoretical results and highlight the impact of a single low-complexity problem early in the sequence.
引用
收藏
页码:599 / 611
页数:13
相关论文
共 50 条
  • [31] Online Distributed Optimization With Nonconvex Objective Functions via Dynamic Regrets
    Lu, Kaihong
    Wang, Long
    IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2023, 68 (11) : 6509 - 6524
  • [32] A Robust Coevolutionary Neural-Based Optimization Algorithm for Constrained Nonconvex Optimization
    Wei, Lin
    Jin, Long
    Luo, Xin
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2024, 35 (06) : 7778 - 7791
  • [33] Constrained robust Bayesian optimization of expensive noisy black-box functions with guaranteed regret bounds
    Kudva, Akshay
    Sorourifar, Farshud
    Paulson, Joel A.
    AICHE JOURNAL, 2022, 68 (12)
  • [34] Distributionally Time-Varying Online Stochastic Optimization under Polyak- Lojasiewicz Condition with Application in Conditional Value-at-Risk Statistical Learning
    Pun, Yuen-Man
    Farokhi, Farhad
    Shames, Iman
    arXiv, 2023,
  • [35] Dynamic Regret Analysis for Event-Triggered Distributed Online Optimization Algorithm
    Yamashita, Makoto
    Hayashi, Naoki
    Takai, Shigemasa
    IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, 2021, E104A (02) : 430 - 437
  • [36] Online Optimization in Dynamic Environments: Improved Regret Rates for Strongly Convex Problems
    Mokhtari, Aryan
    Shahrampour, Shahin
    Jadbabaie, Ali
    Ribeiro, Alejandro
    2016 IEEE 55TH CONFERENCE ON DECISION AND CONTROL (CDC), 2016, : 7195 - 7201
  • [37] Online distributed optimization algorithm with dynamic regret analysis under unbalanced graphs
    Yao, Songquan
    Xie, Siyu
    Li, Tao
    AUTOMATICA, 2025, 174
  • [38] Fast constrained dynamic matrix control algorithm with online optimization
    Peccin V.B.
    Lima D.M.
    Flesch R.C.C.
    Normey-Rico J.E.
    RIAI - Revista Iberoamericana de Automatica e Informatica Industrial, 2022, 19 (03): : 330 - 342
  • [39] Universal Online Convex Optimization With Minimax Optimal Second-Order Dynamic Regret
    Gokcesu, Hakan
    Kozat, Suleyman Serdar
    IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2024, 69 (06) : 3865 - 3880
  • [40] Improved dynamic regret of distributed online multiple Frank-Wolfe convex optimization
    Zhang, Wentao
    Shi, Yang
    Zhang, Baoyong
    Yuan, Deming
    SCIENCE CHINA-INFORMATION SCIENCES, 2024, 67 (11)