2581 valuable arXiv mathematics papers the pipeline has analyzed — each with a short note on what its math gives to machine learning.
The paper identifies an explicit finite-dimensional representation of tree ensembles through leaf-indicator features and argues that, as the number of trees grows, these random features converge to a Gaussian-process model with a favorable…
The paper offers a constructive safety mechanism: impose a control-barrier condition on the reference signal of an existing closed-loop controller, using an explicit model of the complete closed-loop plant-controller dynamics. Instead of…
The paper provides a constructive spectral-stabilization principle for discretizations whose instability is concentrated in a small set of coordinates: identify the localized eigenmode, then blend the original operator with a conservative…
The paper provides an exactly solvable exclusion dynamics in which local particle moves are controlled by a global logarithmic-repulsion factor, together with a free-fermion and determinantal representation of the evolution. Its…
The paper offers a transferable observer-based mechanism for rejecting false-data injections when the attack amplitude may be unbounded but its rate is bounded. Its key machinery is augmentation of the plant with an attack-generating…
The paper identifies a constructive mechanism in which quenched disorder and non-reciprocal coupling generate self-sustained oscillations without external periodic forcing. In the complete-graph mean-field dynamics, antisymmetric…
The paper offers a constructive sensitivity pipeline: perturbations in operating conditions or network parameters are propagated through an operating-point Jacobian and then mapped to shifts in closed-loop eigenvalues. Its transferable…
The paper identifies two highly transferable engineering principles for PINNs: enforce geometric boundary and asymptotic constraints through the network parametrization, and evaluate second-order PDE residuals with forward-propagated…
The paper gives a concrete acceleration strategy for random-sweep Gibbs sampling by moving a Gaussian graphical model into a Fourier-dual normal factor graph. The transferable assets are dual-domain sampling, exact spectral convergence…
The paper's transferable contribution is a smoothing-based variational treatment of discrete graph partitioning: instead of committing to hard communities before representation learning, it samples partition-inducing weights from a…
The paper provides a constructive way to enlarge a locally defined combinatorial object while changing its global connectivity, using finite graph lifts generated by a group extension and a chosen section. This suggests neural…
The paper identifies random graph geometry, rather than random couplings or interactions, as a mechanism that produces sharply different transport regimes in a noninteracting system. Its constructive model is a layered chain whose adjacent…
The paper provides an exact decision-diagram representation of binary multilinear feasible sets and converts it into a network-flow extended formulation whose LP relaxation is integral. This structure is transferable to neural-network…
The paper provides a constructive subspace predictor for affine LPV systems that separates past data, future inputs, scheduling trajectories, and innovations, then compresses the resulting lifted predictor using an LQ factorization. Its…
The paper develops a non-autonomous KAM mechanism for persistence of lower-dimensional isotropic invariant tori when a Hamiltonian system has normally parabolic, rather than normally hyperbolic, transverse dynamics and is subjected to…
The paper provides a constructive boundary scaling law for solutions of nonlocal nonlinear Dirichlet problems: the natural vanishing profile is not generally linear in the distance to the boundary, but ${\rm d}_{\Omega}^{s}$. This suggests…
The paper establishes that the worst-case lattice covering density of an n-dimensional convex body scales as n log n up to iterated-logarithmic factors. This suggests a structured alternative to coordinatewise quantization: learn an affine…
The paper turns gradient-based event simulation into a query-efficient procedure by evaluating the gradient only at deterministic window boundaries and using smoothness to construct an upper envelope for the event intensity inside each…
The paper supplies a concrete multiscale algebra for representing transformations: elements are reduced pairs of rooted binary trees together with a leaf permutation, and redundant refinements are removed by canceling matching carets. This…
The paper gives a sharp negative mechanism: identical periodic jobs sharing an anonymous resource have exactly zero pairwise phase coupling, so contention does not synchronize them. The two-worker return map is the identity, while the…
The paper gives explicit finite-dimensional and infinite-dimensional constructions for twisted commutation, especially the relation TS=qST for a unimodular phase q. Its transferable asset is a sparse, parameter-efficient algebra of shift…
The paper offers a constructive approximate robust NMPC mechanism: replace a long nonlinear optimization horizon with a small quadratic program enforcing an input-to-state control Lyapunov function (ISS-CLF) and a robust control barrier…
The paper identifies boundary tangency as the structural condition that lets a no-flux density evolution be represented by a confined regular Lagrangian flow, while warning that vanishing volume density near a wall can create boundary…
The paper supplies a sharp spectral-shape principle: among origin-symmetric Gaussian domains with fixed Gaussian mass, the centered ball minimizes the reciprocal sum of the first N nonzero Ornstein–Uhlenbeck Neumann eigenvalues. The…