Complexity and Non-Commutativity of Learning Operations on Graphs

H. Atmanspacher; T. Filk

Home/Archive/arXiv

arXiv · nlin/0512055

Complexity and Non-Commutativity of Learning Operations on Graphs

H. Atmanspacher·T. Filk

View original source ↗Read source document ↗Export metadata ↓

Abstract

We present results from numerical studies of supervised learning operations in recurrent networks considered as graphs, leading from a given set of input conditions to predetermined outputs. Graphs that have optimized their output for particular inputs with respect to predetermined outputs are asymptotically stable and can be characterized by attractors which form a representation space for an associative multiplicative structure of input operations. As the mapping from a series of inputs onto a series of such attractors generally depends on the sequence of inputs, this structure is generally non-commutative. Moreover, the size of the set of attractors, indicating the complexity of learning, is found to behave non-monotonically as learning proceeds. A tentative relation between this complexity and the notion of pragmatic information is indicated.

Explore related subjects

nlin.AO nlin.CD

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

H. Atmanspacher, T. Filk. 2005-12-20. Complexity and Non-Commutativity of Learning Operations on Graphs. https://arxiv.org/abs/nlin/0512055

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Inferring Coupled Stuart-Landau Equations from Waveforms

We present a physics-informed, data-driven framework for inferring phase-amplitude equations of coupled limit-cycle oscillators from waveform measurements. Guided by normal-form theory near a supercritical Hopf bifurcation, we infer a near-identity transformation from two observables of an isolated oscillator to a complex-amplitude coordinate, together with the intrinsic Stuart-Landau dynamics. Applying this transformation to coupled oscillators then enables estimation of the leading-order linear coupling terms. Using coupled van der Pol oscillators as an analytical benchmark, we demonstrate accurate inference under system noise and quantify the validity range with respect to the Hopf bifurcation parameter and coupling strength. Moreover, a model inferred from one informative trajectory at each coupling condition reproduces synchronized states and their basin structure for initial conditions not used in the inference. Applied to a high-dimensional hydrodynamic system, the inferred model captures bistability between in-phase and anti-phase synchronization and identifies the loss of anti-phase synchrony as a Neimark-Sacker bifurcation. Our framework provides a practical route from experimentally accessible waveforms to an interpretable normal-form model for stability and bifurcation analysis.

nlin.AO↗

arXiv2026

Fluctuations in Environmental Resource Availability Alter the Oscillatory Tragedy of the Commons

Environmental noise in real-world complex systems--including fluctuations in temperature, humidity, and related environmental conditions--inevitably modulates individual metabolic capacity and, consequently, the efficiency of public-resource utilization. Yet conventional evolutionary game models incorporating environmental feedback treat resource dynamics deterministically. Here, we incorporate stochastic fluctuations into a noise-corrected replicator-dynamics framework. Theoretical analysis and stochastic birth-death simulations yield a counterintuitive result: moderate environmental noise stabilizes, rather than destabilizes, the interior fixed point across diverse payoff landscapes, while suppressing or eliminating catastrophic oscillations. These findings reframe environmental variability not merely as a perturbation but as a mechanism that fosters strategy coexistence and sustainable resource use in complex social and ecological systems.

nlin.AO↗

arXiv2026

Phase Locking in Oscillator Networks with Uncertain and State-Dependent Couplings

Phase locking in oscillator networks depends critically on the interaction strengths between coupled units. In many applications, however, these couplings are uncertain, state dependent, or time varying. We derive rigorous conditions for phase locking in Kuramoto-type networks when each coupling strength is known only within prescribed bounds. Building on a convex formulation of the nonlinear fixed-point equations, we obtain edge-wise error bounds for arbitrary feasible approximations that hold uniformly over all admissible couplings. These bounds yield sufficient conditions for the existence of phase-locked states with prescribed phase-difference limits. We illustrate the framework for AC power grids with uncertain voltage magnitudes and adaptive oscillator networks with state-dependent couplings. For the latter, the bounds can be further tightened by exploiting the state dependence of the couplings.

nlin.AO↗