The adaptive projected subgradient method constrained by families of quasi-nonexpansive mappings and its application to online learning

Konstantinos Slavakis, Isao Yamada

Research output: Contribution to journalArticlepeer-review

17 Scopus citations

Abstract

Many online, i.e., time-adaptive, inverse problems in signal processing and machine learning fall under the wide umbrella of the asymptotic minimization of a sequence of nonnegative, convex, and continuous functions. To incorporate a priori knowledge into the design, the asymptotic minimization task is usually constrained on a fixed closed convex set, which is dictated by the available a priori information. To increase versatility toward the usage of the available information, the present manuscript extends the adaptive projected subgradient method by introducing an algorithmic scheme which incorporates a priori knowledge in the design via a sequence of strongly attracting quasinonexpansive mappings in a real Hilbert space. In such a way, the benefits offered to online learning tasks by the proposed method unfold in two ways: (1) the rich class of quasi-nonexpansive mappings provides a plethora of ways to cast a priori knowledge, and (2) by introducing a sequence of such mappings, the proposed scheme is able to capture the time-varying nature of a priori information. The convergence properties of the algorithm are studied, several special cases of the method with wide applicability are shown, and the potential of the proposed scheme is demonstrated by considering an increasingly important online sparse system/signal recovery task.

Original languageEnglish (US)
Pages (from-to)126-152
Number of pages27
JournalSIAM Journal on Optimization
Volume23
Issue number1
DOIs
StatePublished - May 20 2013

Keywords

  • Adaptive filtering
  • Asymptotic minimization
  • Fixed point
  • Online learning
  • Projection
  • Quasi-nonexpansive mapping
  • Sparsity
  • Subgradient

Fingerprint Dive into the research topics of 'The adaptive projected subgradient method constrained by families of quasi-nonexpansive mappings and its application to online learning'. Together they form a unique fingerprint.

Cite this