Uniform rank gradient, cost, and local-global convergence

DOI10.1090/TRAN/8008MaRDI QIDQ5218237zbMATH OpenOpenAlexFDO

Authors Miklós Abért, László Márton Tóth

Publication date 2 March 2020

Published in Transactions of the American Mathematical Society (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1710.10431

Mathematics Subject Classification ID

Residual properties and generalizations; residually finite groups (20E26) Generators, relations, and presentations of groups (20F05) Asymptotic properties of groups (20F69) General groups of measure-preserving transformations and dynamical systems (37A15) Chains and lattices of subgroups, subnormal subgroups (20E15)

Abstract: We analyze the rank gradient of finitely generated groups with respect to sequences of subgroups of finite index that do not necessarily form a chain, by connecting it to the cost of p.m.p. actions. We generalize several results that were only known for chains before. The connection is made by the notion of local-global convergence. In particular, we show that for a finitely generated group $Gamma$ with fixed price $c$, every Farber sequence has rank gradient $c-1$. By adapting Lackenby's trichotomy theorem to this setting, we also show that in a finitely presented amenable group, every sequence of subgroups with index tending to infinity has vanishing rank gradient.

Recommendations

• Universal gradient methods for convex optimization problems
• scientific article; zbMATH DE number 2196505
• On the Global Convergence of Randomized Coordinate Gradient Descent for Nonconvex Optimization
• Convergence of Gradient Descent for Low-Rank Matrix Approximation
• Uniform approximation: The non-locally convex case
• Fast gradient methods for uniformly convex and weakly smooth problems
• A uniform gradient bound of minimizers of regularized functionals
• scientific article; zbMATH DE number 1778080
• A framework for globally convergent algorithms using gradient bounding functions
• Globally convergent optimization algorithms on Riemannian manifolds: Uniform framework for unconstrained and constrained optimization

Cites work

• Amenability, Kazhdan's property T, strong ergodicity and invariant means for ergodic group-actions
• Bernoulli actions are weakly contained in any free action
• Cost of equivalence relations and groups
• Dynamical properties of profinite actions
• Ergodic theory of amenable group actions. I: The Rohlin lemma
• Expanders, rank and graphs of groups
• Global aspects of ergodic group actions
• Limits of locally-globally convergent graph sequences
• Non-standard limits of graphs and some orbit equivalence invariants
• Rank gradient, cost of groups and the rank versus Heegaard genus problem
• Rank, combinatorial cost, and homology torsion growth in higher rank lattices
• Sparse graphs: metrics and random models
• The combinatoral cost
• The rank gradient from a combinatorial viewpoint.

Cited in

(7)

• Rank gradient and \(p\)-gradient of amalgamated free products and HNN extensions.
• Asymptotic invariants of lattices in locally compact groups
• The rank gradient from a combinatorial viewpoint.
• Point processes, cost, and the growth of rank in locally compact groups
• Non-standard limits of graphs and some orbit equivalence invariants
• Combinatorial cost: a coarse setting
• On a curious problem and what it lead to

This page was built for publication: Uniform rank gradient, cost, and local-global convergence

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5218237)