arXiv research

Constructs perturbations of a minimal surface with triple junctions.

Noncompact Ricci-flat solutions have infinite unstable dimensions.

We extend the results of Hardt and Simon on area-minimizing cones to prove that isolated singularities of stationary one-sided area-minimizing hypersurfaces can be locally perturbed away on the side that they are minimizing.

Two randomized algorithms improve performance in non-stationary linear bandits.

We analyze stochastic gradient algorithms for optimizing nonconvex problems. In particular, our goal is to find local minima (second-order stationary points) instead of just finding first-order stationary points which may be some bad unstable saddle points. We show that a simple perturbed version of stochastic recursiv…

Study on financial systems using perturbed unimodal maps with heteroscedastic noise.

Ricci flow simulations show unstable Fubini-Study metrics develop singularities.

This paper shows that a perturbed form of gradient descent converges to a second-order stationary point in a number iterations which depends only poly-logarithmically on dimension (i.e., it is almost "dimension-free"). The convergence rate of this procedure matches the well-known convergence rate of gradient descent to…

PWGF escapes saddle points in nonconvex optimization.

The alternating gradient descent (AGD) is a simple but popular algorithm which has been applied to problems in optimization, machine learning, data ming, and signal processing, etc. The algorithm updates two blocks of variables in an alternating manner, in which a gradient step is taken on one block, while keeping the …

New method finds stationary points in bilevel optimization problems.

Novel methods for accelerating optimization in complex bilevel and minimax problems.

Study on recovering Lorentzian metrics from scattering data.

Study stationary measures and orbit closures for non-abelian actions on surfaces.

We prove a perturbative result concerning the uniqueness of Kerr-Newman family of black holes: given an asymptotically flat space-time with bifurcate horizons, if it agrees with a non-extremal Kerr-Newman space-time asymptotically flat at infinity and it is sufficiently close to the Kerr-Newman family, then the space-t…

SAMoSSA combines mSSA and AR for accurate time series analysis.

A new fuzzy time series method for non-stationary data.

The paper proves the existence of a horizon for certain spacetimes.

New gauge fields modify Fokker-Planck dynamics without changing the stationary state.

Stability of catenoid in hyperbolic space proven without symmetry assumptions.

We show that a stationary asymptotically flat electro-vacuum solution of Einstein's equations that is everywhere locally "almost isometric" to a Kerr-Newman solution cannot admit more than one event horizon. Axial symmetry is not assumed. In particular this implies that the assumption of a single event horizon in Alexa…

Gradient descent (GD) and stochastic gradient descent (SGD) are the workhorses of large-scale machine learning. While classical theory focused on analyzing the performance of these methods in convex optimization problems, the most notable successes in machine learning have involved nonconvex optimization, and a gap has…

SAM optimizer struggles to converge to global minima or stationary points in practical settings.

In this note we show that the recent dynamical stability result for small $C^1$-perturbations of strongly stable minimal submanifolds of C.-J. Tsai and M.-T. Wang directly extends to the enhanced Brakke flows of Ilmanen. We illustrate applications of this result, including a local uniqueness statement for strongly stab…

We review the spectral analysis and the time-dependent approach of scattering theory for manifolds with asymptotically cylindrical ends. For the spectral analysis, higher order resolvent estimates are obtained via Mourre theory for both short-range and long-range behaviors of the metric and the perturbation at infinity…

The minority game (MG) model introduced recently provides promising insights into the understanding of the evolution of prices, indices and rates in the financial markets. In this paper we perform a time series analysis of the model employing tools from statistics, dynamical systems theory and stochastic processes. Usi…

This paper tackles gauge fixing and regularity for perturbations around spherical backgrounds.

Proves existence and uniqueness of rotating fluid bodies in GR to second order.

Simple gradient descent algorithm escapes saddle points efficiently.

The large $k$ asymptotics (perturbation series) for integrals of the form $\int_{\cal F}μe^{i k S}$, where $μ$ is a smooth top form and $S$ is a smooth function on a manifold ${\cal F}$, both of which are invariant under the action of a symmetry group ${\cal G}$, may be computed using the stationary phase approximation…

Paper introduces metrics for evaluating multi-agent policies using best response dynamics.

Mutation improves FTRL convergence in zero-sum games.

We consider streaming principal component analysis when the stochastic data-generating model is subject to perturbations. While existing models assume a fixed covariance, we adopt a robust perspective where the covariance matrix belongs to a temporal uncertainty set. Under this setting, we provide fundamental limits on…

Designing deterministic denominators for SGLD stabilizes large drifts.

Efficiently samples multimodal distributions using data-based initialization.

We study the Stochastic Gradient Descent (SGD) method in nonconvex optimization problems from the point of view of approximating diffusion processes. We prove rigorously that the diffusion process can approximate the SGD algorithm weakly using the weak form of master equation for probability evolution. In the small ste…

SGD converges to global minimum for certain non-convex functions.

We present a novel approach for learning an HMM whose outputs are distributed according to a parametric family. This is done by {\em decoupling} the learning task into two steps: first estimating the output parameters, and then estimating the hidden states transition probabilities. The first step is accomplished by fit…

In this paper, we consider efficient differentially private empirical risk minimization from the viewpoint of optimization algorithms. For strongly convex and smooth objectives, we prove that gradient descent with output perturbation not only achieves nearly optimal utility, but also significantly improves the running …

Training an artificial neural network involves an optimization process over the landscape defined by the cost (loss) as a function of the network parameters. We explore these landscapes using optimisation tools developed for potential energy landscapes in molecular science. The number of local minima and transition sta…

If a differential equation in a Banach manifold is invariant or quasi-invariant under the action of one or more Lie groups, then its stationary points cannot be isolated, so that classical linearized stability theorem does not apply to it. The first main purpose of this paper is to establish a linearized stability theo…

Analyzes learning dynamics of RNNs under locality constraints.

New algorithm helps escape saddle points in optimization problems.

WassersteinGrad improves weather forecasting explanations by addressing geometric misalignment issues.

We consider minimizing a nonconvex, smooth function $f$ on a Riemannian manifold $\mathcal{M}$. We show that a perturbed version of Riemannian gradient descent algorithm converges to a second-order stationary point (and hence is able to escape saddle points on the manifold). The rate of convergence depends as $1/ε^2$ o…

Maximum likelihood estimation fails to be well-posed in Gaussian process regression.

We address the problem of estimating the mixing time $t_{\mathsf{mix}}$ of an arbitrary ergodic finite-state Markov chain from a single trajectory of length $m$. The reversible case was addressed by Hsu et al. [2019], who left the general case as an open problem. In the reversible case, the analysis is greatly facilita…

Periodic activation functions improve neural network reliability and interpretability.