New metric assesses image dataset complexity.
problem Assessing the complexity of image classification datasets.
method Cumulative Spectral Gradient (CSG) derived from spectral clustering.
result CSG correlates with CNN test accuracy and dataset separability.
Algorithm extracts deterministic PDFA from probabilistic models with improved performance.
problem Learning deterministic models from probabilistic ones with noise.
method Adapted L* algorithm for probabilistic settings, using conditional probabilities and local tolerance.
result Achieves better performance on WER and NDCG than spectral extraction of WFAs.
Boost-R uses gradient boosted trees for analyzing recurrence data.
problem Analyzing recurrence data with static and dynamic features.
method Gradient boosted additive trees with time-dependent functions.
result Estimates the cumulative intensity function of recurrent event processes.
Develops uniform convergence guarantees for a broad class of risk functionals in supervised learning.
problem Bounding generalization gaps for various risk functionals beyond the expectation.
method Establishes uniform convergence for Hölder risk functionals, providing guarantees for empirical risk minimization.
result First uniform convergence results for estimating the CDF of loss distributions, applicable to various risk functionals.
A large class of machine learning techniques requires the solution of optimization problems involving spectral functions of parametric matrices, e.g. log-determinant and nuclear norm. Unfortunately, computing the gradient of a spectral function is generally of cubic complexity, as such gradient descent methods are rath…
The study examines neural networks with random weights and biases, finding that depth-to-width ratio controls fluctuations and correlations.
problem Exploring the exploding and vanishing gradient problem in neural networks with random weights and biases.
method Sharp estimates of joint cumulants and solving cumulant recursions in powers of 1/n.
result The depth-to-width ratio L/n plays a crucial role in controlling fluctuations and correlations, leading to the occurrence of exploding and vanishing gradients. Study spectral density of neural networks using resolvent method.
problem Investigate spectral density of neural networks with random feature matrices.
method Use resolvent method from random matrix theory, cumulant expansion.
result Impossible to preserve singular value distribution with additive bias.
New method for RL with general utilities using variational policy gradient.
problem Optimizing policies with general concave utility functions in RL.
method Derives Variational Policy Gradient Theorem, develops variational Monte Carlo gradient estimation algorithm.
result Global convergence to optimal policy for general objectives, exponential convergence under strong convexity.
A new GAN loss function based on cumulant generating functions improves stability and robustness.
problem Improving the stability and performance of GANs.
method Cumulant GAN loss function based on variational R{é}nyi divergence.
result Cumulant GAN achieves linear convergence to Nash equilibrium and superior performance in image generation.
Spectral gradient methods outperform Euclidean in certain deep learning scenarios.
problem When do spectral gradient updates outperform Euclidean in deep learning?
method Layerwise condition comparing squared nuclear-to-Frobenius ratio to stable rank of activations.
result Spectral updates can be more effective than Euclidean in deep networks and transformers.
In this paper we consider dividend problem for an insurance company whose risk evolves as a spectrally negative Lévy process (in the absence of dividend payments) when Parisian delay is applied. The objective function is given by the cumulative discounted dividends received until the moment of ruin when so-called barri…
A new algorithm learns optimal source placement in large networks.
problem Optimizing source placement in large scale networks with unknown processes.
method Graph-Kernel Multi-Armed Bandit (Grab-UCB) algorithm with adaptive graph dictionary model.
result Online learning algorithm outperforms offline methods in terms of cumulative regret, sample efficiency, and computational complexity.
New method improves GAN training stability and quality.
problem Improving training stability and sample quality in GANs.
method Proposes a new method for Lipschitz continuity in GANs that is efficient and unbiased.
result Demonstrates the effectiveness of the new method in various GAN training scenarios.
Spectral normalization stabilizes GANs by controlling gradient explosion and vanishing.
problem Stability and sample quality issues in GAN training.
method Spectral normalization controls gradient explosion and vanishing, improving GAN training stability and sample quality.
result Bidirectional Scaled Spectral Normalization (BSSN) outperforms standard spectral normalization in sample quality and training stability.
Spectral deconfounding improves machine learning models by reducing hidden confounding effects.
problem Machine learning models can be misled by hidden confounders, leading to unreliable predictions.
method Develops a nonlinear spectral deconfounding framework for gradient boosting that modifies boosting dynamics to slow down in confounding-aligned directions.
result Spectrally deconfounded boosting improves estimation of the target function under hidden confounding and is more scalable.
Unified approach to tensor PCA and related problems using tensor cumulants.
problem Statistical inference on invariant distributions, particularly tensor PCA.
method Definition and analysis of tensor cumulants to unify and extend previous results.
result Unified explanation of hardness and subexponential-time algorithms for tensor PCA.
SpecGD mitigates misalignment in phase retrieval models with anisotropic inputs.
problem Misalignment during gradient descent in phase retrieval models with anisotropic inputs.
method Spectral gradient descent modifies gradient updates to preserve directional information and remove spike amplification.
result SpecGD removes spike amplification, leading to stable alignment and accelerated noise contraction.
New construction of Fukaya-Seidel categories using complex gradient flow equation.
problem Constructing Fukaya-Seidel categories for specific models.
method Using the complex gradient flow equation and neck-stretching limits.
result Alternative proof of Seidel's spectral sequence for Lagrangian Floer cohomology.
This paper studies the optimal dividend problem with capital injection under the constraint that the cumulative dividend strategy is absolutely continuous. We consider an open problem of the general spectrally negative case and derive the optimal solution explicitly using the fluctuation identities of the refracted-ref…
Gradient descent with small random init mimics spectral methods for low-rank matrix recovery.
problem Reconstructing a low-rank matrix from few measurements.
method Gradient descent with small random initialization followed by a few iterations.
result Gradient descent from small random init converges to a well-generalizing solution.
The objective in a traditional reinforcement learning (RL) problem is to find a policy that optimizes the expected value of a performance metric such as the infinite-horizon cumulative discounted or long-run average cost/reward. In practice, optimizing the expected value alone may not be satisfactory, in that it may be…
Neural networks learn faster with correlated latent variables.
problem Efficiently learning from higher-order correlations in neural networks.
method Analytical derivation and simulations of two-layer neural networks.
result Correlations between latent variables speed up learning from higher-order correlations.
Method detects neural network equivalence via matrix ensembles and spectral analysis.
problem Detecting equivalence among different deep learning architectures.
method Generating Mixed Matrix Ensembles (MMEs) and matching to conjugate circular ensembles.
result Empirical evidence shows vanishing differences in spectral densities with long tail decay rates.
Study of correlated Wigner matrices with BBP transitions.
problem Understanding spectral transitions in correlated Wigner matrices.
method Analyzes a Wigner-type matrix with row/column correlations, decomposes into bulk and outliers, and uses integral operators to model transitions.
result Correlated Wigner matrices exhibit multiple BBP transitions at critical points.
Spectral portfolio theory links neural networks to wealth dynamics via SGD weight matrices.
problem Understanding wealth dynamics from neural network training.
method Direct identification of weight matrices as portfolio allocation matrices, linking SGD forces to portfolio dynamics.
result Spectral properties of SGD weight matrices transition between additive and multiplicative regimes, influencing wealth dynamics.
Improves learning of spectral mixture kernels with approximate Bayesian inference.
problem Difficult optimization of large number of SM kernel parameters.
method Approximate Bayesian inference using variational distribution of spectral points and random Fourier features.
result Accelerates convergence and leads to better optimal parameters.
Muon optimizer simplifies matrix optimization with spectral orthogonalization.
problem Matrix optimization challenges, especially with large condition numbers.
method Simplified Muon optimizer using spectral orthogonalization of gradients.
result Simplified Muon converges linearly with independent scalar sequences, outperforming gradient descent and Adam.
New method accelerates smooth games using spectral shape analysis.
problem Accelerating optimization in smooth games with complex numerical challenges.
method Matrix iteration theory and spectral shape analysis to characterize and manipulate acceleration.
result Identified a continuum of optimization strategies from convex minimization to gradient descent.
In this paper we study a spectrally negative Lévy process which is refracted at its running maximum and at the same time reflected from below at a certain level. Such a process can for instance be used to model an insurance surplus process subject to tax payments according to a loss-carry-forward scheme together with t…
The paper analyzes the variance of different shuffling methods in stochastic gradient descent.
problem Understanding the variance of different shuffling methods in stochastic gradient descent.
method Power spectral density analysis to study the noise sequences of stochastic gradients.
result The stationary variances of iterates decrease in the order of SGD, SGD-RR, and SGD-SO.
Optimistic algorithm reduces regret and constraint violations in online convex optimization with adversarial constraints.
problem Online convex optimization with adversarial constraints.
method Improved algorithm using accurate predictions of loss and constraint functions.
result Improved bounds on regret and cumulative constraint violations.
Proposes a new adaptive gradient method based on gradient differences.
problem Manual tuning of stepsize in vanilla gradient methods.
method Adaptation driven by cumulative squared norms of gradient differences.
result More robust than AdaGrad in various settings.
Optimal dividend strategy found for risk models with regime switching.
problem Optimal dividend strategy for spectrally negative Markov additive models with regime switching.
method Introduced an auxiliary problem and transformed the original problem into a local optimization problem.
result The refraction-reflection strategy with regime-modulated thresholds is optimal.
Muon dynamics study uses spectral Wasserstein flow for optimization stability.
problem Optimizing deep learning models with gradient normalization.
method Introduces Spectral Wasserstein distances for matrix flows, proving equivalence with Benamou--Brenier formulation.
result Gradient-flow interpretation of mean-field normalized training dynamics.
The paper examines how gradient descent stabilizes low-rank matrix factorization in noisy conditions.
problem Stability of low-rank implicit regularization in perturbed deep matrix factorization.
method Derives spectral conditions for gradient descent to exhibit a low-rank phase in noiseless settings and analyzes perturbed dynamics.
result Gradient descent converges to a low-rank solution under perturbation, with explicit dependence on perturbation size.
Deep networks learn clean structure before memorizing corrupted labels, leaving a spectral signature in gradient centered scatter.
problem Deep networks' transition from learning clean structure to memorizing corrupted labels under label noise.
method Analysis of the centered scatter of per-example last-layer gradients to identify Fisher Rank Inflation.
result Fisher Rank Inflation is a spectral signature of memorization under label noise, with effective rank expanding during memorization.
Investigates spectral properties of neural networks, showing invariance under certain conditions.
problem Understanding the spectral evolution and invariance in linear-width neural networks.
method Empirical and theoretical analysis of spectra of weight matrices in high-dimensional settings.
result Spectra of weight matrices are invariant under certain training conditions, with implications for feature learning.
The paper proposes a new model for financial order books without assuming prices or quantities.
problem Understanding the geometry of financial order books without assuming prices or quantities.
method Modeling financial order books as an inflationary relational system without metric, temporal, or price coordinates. Observable quantities arise through spectral embeddings of the graph Laplacian.
result Projected supply and demand are constrained to gamma-like functional forms, which can be observed as integrated-gamma cumulative profiles in high-frequency data.
Study f-Laplace bounds on gradient Ricci shrinkers, applying to Betti numbers.
problem Bounding eigenvalues of f-Laplacian on gradient Ricci shrinkers. method Upper and lower bounds established using volume growth rate; extends to vector bundles.
result Explicit upper bounds for Betti numbers derived.
This work analyzes a two-stage algorithm for single index models, showing precise asymptotics of gradient descent.
problem Learning single index models with non-convex optimization.
method Spectral initialization followed by gradient descent, with detailed analysis of dynamics and asymptotics.
result Gradient descent converges to long-time fixed points in the large system limit, representing mean field behavior.
Proposes MSN to improve DNN performance and speed.
problem Improving Deep Learning model regularization and performance.
method Empirical approach to study Spectral Normalization (SN) and Mean Spectral Normalization (MSN).
result MSN significantly improves DNN performance and speed.
In this paper, we propose a scalable algorithm for spectral embedding. The latter is a standard tool for graph clustering. However, its computational bottleneck is the eigendecomposition of the graph Laplacian matrix, which prevents its application to large-scale graphs. Our contribution consists of reformulating spect…
Sharp pseudospectral bounds prevent transient amplification in coupled gradient descent.
problem Transient amplification in coupled gradient descent systems.
method Developed a sharp pseudospectral theory for block-triangular Jacobians, proving Kreiss constant bounds and matching minimax lower bounds.
result Obtained a finite-horizon iteration-complexity bound of O(K(J)2log(1/δ)) for stochastic coupled descent. SpecNet2 improves spectral embedding without orthogonalization, achieving better performance and efficiency.
problem Improving spectral embedding methods for better performance and efficiency.
method Optimizes an equivalent objective of the eigen-problem without orthogonalization, allowing separate row and column sampling.
result Local and global convergence of the new objective using batch-based gradient descent is proven, and improved performance and efficiency are demonstrated on simulated and image datasets.
Over the past decade there has been considerable interest in spectral algorithms for learning Predictive State Representations (PSRs). Spectral algorithms have appealing theoretical guarantees; however, the resulting models do not always perform well on inference tasks in practice. One reason for this behavior is the m…
In this paper we introduce a new coherent cumulative risk measure on RLp, the space of càdlàg processes having Laplace transform. This new coherent risk measure turns out to be tractable enough within a class of models where the aggregate claims is driven by a spectrally positive Lévy process. Moreover, w…
Simple proof that stable minimal hypersurfaces in R^4 are hyperplanes.
problem Proving stable minimal hypersurfaces in R^4 are hyperplanes.
method Using spectral Ricci curvature bounds and Green kernel estimates.
result Complete, two-sided stable minimal hypersurfaces in R^4 are hyperplanes.
Unified probabilistic gradient boosting for entire conditional distribution modeling.
problem Creating accurate probabilistic forecasts from regression tasks.
method Unified probabilistic gradient boosting framework using XGBoost and LightGBM, modeling conditional moments or CDF via Normalizing Flows.
result Achieves state-of-the-art forecast accuracy.