Quantum algorithm speeds up MIP solving by a near-quadratic factor.
problem Solving Mixed Integer Programs (MIPs) efficiently.
method Incremental-Quantum-Branch-and-Bound algorithm combining quantum speedup with classical search heuristics.
result Universal near-quadratic speedup over classical Branch-and-Bound algorithms.
New method makes robust estimators work without knowing corruption levels.
problem Robust estimation algorithms struggle with unknown corruption levels.
method Abstracted geometric puzzle solution to universal meta technique.
result Converts any robust estimator to work without corruption bounds.
Under-parameterized networks can either copy or average teacher weights, leading to universal optimal solutions.
problem Approximating a teacher network with an under-parameterized student network.
method Analyzing shallow neural networks with erf activation function and unitary teacher weights, proving copy-average configurations are critical points and finding the optimal solution.
result The optimal solution for under-parameterized networks has a universal structure, whether copying or averaging teacher neurons.
Simple construction for universal quantum gates.
problem Designing efficient quantum gates for topological computers.
method Demonstrated a simple construction for unitary solutions of the braided Yang-Baxter equation in any dimension.
result Proved the existence of universal quantum gates in any dimension.
FLASH-MAX predicts electromagnetic fields from sparse data in seconds.
problem Predicting homogeneous electromagnetic fields from sparse pointwise observations.
method Exact-by-construction neural network architecture that satisfies Maxwell's equations symbolically.
result FLASH-MAX achieves sub-1% relative validation error from 1K sparse observations in seconds.
This paper considers the design of optimal resource allocation policies in wireless communication systems which are generically modeled as a functional optimization problem with stochastic constraints. These optimization problems have the structure of a learning problem in which the statistical loss appears as a constr…
We establish blow-up profiles for any blowing-up sequence of solutions of general conformally invariant fully nonlinear elliptic equations on Euclidean domains. We prove that (i) the distance between blow-up points is bounded from below by a universal positive number, (ii) the solutions are very close to a single stand…
Although stochastic gradient descent (SGD) method and its variants (e.g., stochastic momentum methods, AdaGrad) are the choice of algorithms for solving non-convex problems (especially deep learning), there still remain big gaps between the theory and the practice with many questions unresolved. For example, there is s…
Universal geometry organizes heterotic moduli spaces.
problem Understanding the structure of heterotic compactifications.
method Fibering compactification data over moduli space and analyzing universal curvatures.
result Deformations are components of universal curvatures, incorporating α′2 corrections. Soft-Radial Projection solves gradient saturation in constrained deep learning.
problem Gradient saturation in deep learning models when integrating hard constraints.
method Introduces Soft-Radial Projection, a differentiable layer that maps predictions onto constraint boundaries without rank-deficient Jacobians.
result Improves convergence and solution quality over state-of-the-art methods.
Dubrovin duality connects two F-manifolds on the universal curve.
problem Connecting two F-manifolds on the universal curve.
method Proving natural extension of Dubrovin dual to F-manifolds with compatible flat connection.
result Equips the universal curve with two F-manifolds with compatible flat structure.
The quantum navigation problem of finding the time-optimal control Hamiltonian that transports a given initial state to a target state through quantum wind, that is, under the influence of external fields or potentials, is analysed. By lifting the problem from the state space to the space of unitary gates realising the…
Develops a universal waveform selection scheme for radar tracking.
problem Optimal waveform selection for target tracking in active sensors.
method Uses reinforcement learning and universal source coding techniques.
result Achieves optimal waveform selection for any radar scene modeled as a Markov process.
Paper interprets ResNets via gate-network controls and deep-layer classifications.
problem Understanding the performance mechanism of ResNets.
method Constructs typical solutions using gate-network controls and deep-layer classifications.
result Proves the universal-approximation capability of ResNets.
Universal online optimization for dynamic environments using uniclass prediction.
problem Online optimization in changing environments with dynamic regret.
method Reduces dynamic online optimization to uniclass prediction problem, allowing control over dynamic regret bounds.
result First paper with state-of-the-art dynamic regret guarantees for general convex cost functions.
The paper finds exact solutions to a complex Einstein-Dirac-Maxwell system on 4D Sasakian spacetimes.
problem Finding exact solutions to an Einstein-Dirac-Maxwell system with Sasakian quasi-Killing spinors.
method Constructing a family of exact solutions on four-dimensional static Sasakian spacetimes using the Sasakian frame.
result Closed and open universe models are found with specific energy conditions.
We calculate the optimal solutions of the fully heterogeneous Von Neumann expansion problem with N processes and P goods in the limit N→∞. This model provides an elementary description of the growth of a production economy in the long run. The system turns from a contracting to an expanding phase as N in…
Proposes autoencoding with random forests using spectral graph theory.
problem Learning low-dimensional embeddings of random forest models.
method Combines nonparametric statistics and spectral graph theory for optimization.
result Establishes a universal consistent decoder for random forest models.
Study of Hitchin moduli spaces over Teichmüller space.
problem Metric aspects of Hitchin moduli spaces over varying complex structures.
method Gauge theoretical approach, Kähler fibrations, moment map interpretation, symplectic reduction.
result Establishes natural complex and pseudo-Kähler structures on universal Hitchin moduli spaces.
Paper constructs solutions to WDVV equations for Frobenius manifolds.
problem Solving open WDVV equations for Frobenius manifolds.
method Explicit construction of flat F-manifolds and principal hierarchies.
result Recovery of polynomial solutions for A- and D-type singularities.
Universal approximation for stochastic processes using Brownian motion.
problem Approximating stochastic processes with linear functionals.
method Establishing Lp-type universal approximation theorems for rough path spaces. result Linear functionals on the signature of time-extended Brownian motion can approximate any p-integrable stochastic process. Study of Gödel Universe as Lie group with specific metric.
problem Characterize geodesics in the Gödel Universe.
method Geometric theory of optimal control applied to Lie groups with left-invariant Lorentz metrics.
result No closed timelike or isotropic geodesics in the Gödel Universe.
Theory extends optimal learning rates without realizability assumption.
problem Agnostic binary classification without realizability assumption.
method Identifies tetrachotomy of optimal rates and combinatorial structures.
result Optimal universal rates for binary classification in agnostic setting.
New framework for conditional risk minimization using optimal transport.
problem High-stakes decisions with side information, especially economic conditions.
method Universal framework based on union-ball formulation in optimal transport.
result Offers interpretability, tractability, and scalability for various risk functionals.
Simulated annealing is a popular method for approaching the solution of a global optimization problem. Existing results on its performance apply to discrete combinatorial optimization where the optimization variables can assume only a finite set of possible values. We introduce a new general formulation of simulated an…
New method achieves both universality and adaptivity in online convex optimization.
problem Achieve optimal regret guarantees without prior knowledge of function curvature.
method Introduces UniGrad, a novel approach that achieves both universality and adaptivity.
result Achieves universal regret guarantees that adapt to gradient variation.
The Navier-Stokes equations on certain manifolds can perform universal computation.
problem Computational universality in viscous fluids.
method Cosymplectic geometry and harmonic 1-forms.
result Stationary Navier-Stokes solutions exhibit Turing completeness.
In this paper we prove that all initially-smooth solutions of the Euler-Weil-Petersson equation, which describes geodesics on the universal Teichmüller space under the Weil-Petersson metric, will remain smooth for all time. This extends the work of Escher-Kolev for strong Riemannian metrics to the borderline case of $H…
We consider the problem of mean-variance portfolio optimization for a generic covariance matrix subject to the budget constraint and the constraint for the expected return, with the application of the replica method borrowed from the statistical physics of disordered systems. We find that the replica symmetry of the so…
In this paper, we investigate the capability of the universal Kriging (UK) model for single-objective global optimization applied within an efficient global optimization (EGO) framework. We implemented this combined UK-EGO framework and studied four variants of the UK methods, that is, a UK with a first-order polynomia…
New approach tackles decision-making under predictions that shape outcomes.
problem Challenges in learning optimal decision rules when predictions influence outcomes.
method Introduces performative omniprediction, a predictor that encodes optimal decision rules for multiple objectives.
result Efficient performative omnipredictors exist under a natural restriction of outcome performativity.
Spectral algorithms solve optimal community detection and related problems.
problem Optimal detection of community structures and related substructures.
method Spectral algorithms applied to various planted substructures.
result Spectral algorithms achieve optimal performance for a wide range of planted substructures.
Researchers found the longest arcs for specific sub-Lorentzian structures.
problem Finding the longest arcs for sub-Lorentzian structures.
method Optimal control problem with unbounded control set and concave cost functional. Sufficient conditions for existence of longest arcs proposed.
result Existence of the longest arcs for left-invariant three-dimensional contact sub-Lorentzian structures proved.
MixMax improves model performance across different settings using convex optimization.
problem Worst-case performance in group distributionally robust optimization for non-convex and non-parametric models.
method Reparameterizing group DRO from parameter space to function space, resulting in a convex optimization problem.
result MixMax matches or outperforms standard group DRO baselines, improving XGBoost performance on specific datasets.
The book explains deep learning theory and how networks learn nontrivial representations.
problem Understanding and optimizing deep neural networks.
method Developed RG flow to characterize signal propagation, solved layer-to-layer equations, and analyzed representation learning.
result Predictions of trained networks are nearly-Gaussian, with depth-to-width ratio controlling deviations.
Streets and Tian introduced a parabolic flow of pluriclosed metrics. We classify the long time behavior of homogeneous solutions of this flow on closed complex surfaces including minimal Hopf, Inoue, Kodaira, and non-Kahler, properly elliptic surfaces. We also construct expanding soliton solutions to the flow on the un…
The paper optimizes regret using covariance between costs and decisions.
problem Optimizing expected regret in decision-making problems.
method Developed derivative theory of covariance regret functional, derived Gâteaux derivative, and extended to constrained optimization.
result Gradient of covariance regret is the cost covariance matrix, with implications for portfolio optimization.
Market sectors play a key role in the efficient flow of capital through the modern Global economy. We analyze existing sectorization heuristics, and observe that the most popular - the GICS (which informs the S&P 500), and the NAICS (published by the U.S. Government) - are not entirely quantitatively driven, but rather…
Study shows the second fundamental form of pseudospherical surfaces is universal and not dependent on specific solutions.
problem Dependence of the second fundamental form in local isometric immersions of pseudospherical surfaces.
method Analysis of third order differential equations and jets of finite order.
result The second fundamental form of pseudospherical surfaces is universal and not dependent on the specific solution.
Survey on computational models in dynamical systems, including new universality concepts.
problem Understanding the relationship between computational models and dynamical systems.
method Review of recent works on Turing universality, Topological Kleene Field Theories, and dynamical bordisms.
result Introduction of new perspectives on computability through dynamical systems.
In this paper, we study adaptive online convex optimization, and aim to design a universal algorithm that achieves optimal regret bounds for multiple common types of loss functions. Existing universal methods are limited in the sense that they are optimal for only a subclass of loss functions. To address this limitatio…
Constructs universal local deformations for curves and differential forms.
problem Local deformations of curves and differential forms under preservation of periods.
method Develops Kuranishi families for pairs of curves and meromorphic 1-forms, focusing on hyperelliptic cases.
result First paper in a series developing a deformation theory for spectral curve data of integrable systems.
Sumformer simplifies Transformers to handle long sequences efficiently.
problem Quadratic complexity of Transformers limits their use with long sequences.
method Introducing Sumformer, a simple architecture that universally approximates equivariant sequence-to-sequence functions.
result Sumformer achieves the first universal approximation results for Linformer and Performer.
A new kernel for probability measures based on optimal transport.
problem Efficiently comparing and modeling distributions.
method Kernel over probability measures using regularized optimal transport and Hilbertian embedding.
result The proposed kernel enables Gaussian process modeling on distributions with theoretical and computational advantages.
A scalable gradient-based framework for sparse portfolio selection.
problem Sparse minimum-variance portfolio selection with cardinality constraint.
method Gradient-based optimization with Boolean relaxation and tunable parameter.
result Matches commercial solvers in most instances, differing by a few assets with negligible error in portfolio variance.
If g(t) is a three-dimensional Ricci flow solution, with sectional curvatures that decay like the inverse of t and diameter that increases at most like the square root of t, then the pullback Ricci flow solution on the universal cover approaches a homogeneous expanding soliton.
We study the problem of learning to rank from multiple information sources. Though multi-view learning and learning to rank have been studied extensively leading to a wide range of applications, multi-view learning to rank as a synergy of both topics has received little attention. The aim of the paper is to propose a c…
Lecture notes on linear neural networks for deep learning optimization and generalization.
problem Understanding optimization and generalization in deep learning models.
method Mathematical tools and dynamical systems theory.
result Potential of mathematical tools to enhance understanding of deep learning.