archive
Every paper Pith has read. Search by title, abstract, or pith.
2292 papers in math.OC · page 9
-
Frugal algorithm achieves zero regret at O(log T) movement
Convex Optimization with Nested Evolving Feasible Sets
-
Co-contraction tunes aerodynamic damping in dual-rotor actuators at fixed thrust
Variable Aerodynamic Damping via Co-Contraction: A Dynamic Isomorphism with Variable Stiffness Actuators
-
Integer vars need Ω(1/ε²) samples for smooth convex stochastic opt
Sample Complexity of Stochastic Optimization with Integer Variables
-
Faster pursuer captures evader in finite time on ellipse
Dynamical Systems in Elliptical Pursuit and Evasion
-
Faster pursuer captures evader in finite time on ellipse
Dynamical Systems in Elliptical Pursuit and Evasion
-
Coderivative conditions ensure stability of conic Nash equilibria
Stability of Lagrangian Generalized Nash Equilibriums
-
Prox-PEP hits stationarity at O(T^{-1/4}) rate for weakly convex stochastic problems
Prox-PEP: A Proximal Partial Exact Penalty Algorithm for Weakly Convex Stochastic Nonlinear Programming
-
Symplectic method reduces quantum models while preserving physics
Symplectic H2 Model Reduction for High-Dimensional Linear Quantum Systems
-
Neural shift operators stabilize HJB policy evaluation
Stabilized neural Hamilton--Jacobi--Bellman solvers: Error analysis and applications in model-based reinforcement learning
-
GNN proxy speeds exact Max-Cut bounding up to 10.6 times
Solving Max-Cut to Global Optimality via Feasibility-Preserving Graph Neural Networks
-
Poisson-Moreau drift yields near-optimal almost sure rates for Markovian SA
Almost Sure Convergence Rates of Stochastic Approximation and Reinforcement Learning via a Poisson-Moreau Drift
-
LS-SVM reformulation enables CV-based feature selection for SVMs
Cross-validation-based optimal feature selection for linear SVM classification
-
Notes present convex optimization theory via first-order methods
Lectures on optimization
-
Newton steps for constrained control reduce to reweighted Riccati solves
A Semi-smooth Newton Method for the Constrained Optimal Control of Continuous-Time Linear Systems
-
Dynamic prices cut synchronized peaks from automated homes
Unlocking Deep Demand Flexibility via Dynamic Signals
-
Relaxation makes dynamic auto-insurance pricing asymptotically optimal at scale
Prescriptive Optimization for Adaptive Auto-insurance Pricing with Telematics Data
-
Auxiliary loss replaces squared gradients with cheap Hessian approximations
Low-Order Explicit Hessian Imitation Method for Large-Scale Supervised Machine Learning
-
Foundation model learns pricing from simulated choices
Causal-Aware Foundation-Model for Bilevel Optimization in Discrete Choice Settings
-
Muon attains optimal rates for heavy-tailed matrix optimization
Muon with Nesterov Momentum: Heavy-Tailed Noise and (Randomized) Inexact Polar Decomposition
-
Port-Hamiltonian optimizer escapes local minima via event triggers
When Descent Is Too Stable: Event-Triggered Hamiltonian Learning to Optimize
-
Embeddings turn async categorical TD into contracting recursions
A Finite-Iteration Theory for Asynchronous Categorical Distributional Temporal-Difference Learning
-
Generic atoms yield simplicial faces in the pseudo-moment cone
Simplicial Regularizability of the Pseudo-Moment Cone and Carath\'eodory-Type Atomic Decomposition of Moment Matrices
-
Rod flow tracks Adam at edge of stability better than stable flows
A Rod Flow Model for Adam at the Edge of Stability
-
Global LLM rankings cancel out two-thirds of votes
Why Global LLM Leaderboards Are Misleading: Small Portfolios for Heterogeneous Supervised ML
-
Matching pretraining optimizer cuts forgetting in full finetuning
Optimizer-Model Consistency: Full Finetuning with the Same Optimizer as Pretraining Forgets Less
-
Multiple trust regions improve high-dimensional Bayesian optimization
MTRBO: Multiple trust-region based Bayesian optimization
-
SignSGD outperforms SGD by factor of d under sparse noise
When and Why SignSGD Outperforms SGD: A Theoretical Study Based on $\ell_1$-norm Lower Bounds
-
Positivity estimate yields classical master equations on graphs without boundaries
Master equations with an individual noise on finite state graphs
-
-
One adjoint pass yields policy gradients plus all input sensitivities
SNAPO: Smooth Neural Adjoint Policy Optimization for Optimal Control via Differentiable Simulation
-
Single-query ZO reaches Goldstein points at O(d² δ^{-3} ε^{-3}) cost
Stochastic Non-Smooth Non-Convex Optimization with Decision-Dependent Distributions
-
Reward early stops to make anytime tests time-sensitive
Time-sensitive anytime-valid testing
-
RL learns to pick cuts that speed Benders decomposition
Learning to Cut: Reinforcement Learning for Benders Decomposition
-
Vectorized formulation linearizes causality for batch SOC
Global self-optimizing control of batch processes
-
Adversary injection makes unobservable quadratic systems observable
Confidentiality of Linear Control Systems with Quadratic Output Under Sensor Attacks [Extended Version]
-
Injected signal restores observability in quadratic-output systems
Confidentiality of Linear Control Systems with Quadratic Output Under Sensor Attacks [Extended Version]
-
Holding learned variables constant optimizes variable-horizon processes
Dynamic Controlled Variables Based Dynamic Self-Optimizing Control
-
Cost-guided training tightens MPC policy optimality bounds
Performance guaranteed MPC Policy Approximation via Cost Guided Learning
-
Second-order bilevel method reaches Õ(ε^{-1.5}) iterations
Second-Order Bilevel Optimization with Accelerated Convergence Rates
-
Gaussian targets make unbalanced transport convex for linear systems
Unbalanced Optimal Transport and Density Control for Discrete-Time Linear Systems
-
Decomposition strategies boost LLM heuristics for coupled optimization
CoupleEvo: Evolving Heuristics for Coupled Optimization Problems Using Large Language Models
-
First-order method converges on bilevel problems over changing directed networks
FAB: A First-Order AB-based Gradient Algorithm for Distributed Bilevel Optimization over Time-Varying Directed Graphs
-
PACE shrinks ensembles by generating then pruning learners
PACE: Prune-And-Compress Ensemble Models
-
Optimality conditions derived for jump diffusions with threshold discontinuities
Stochastic Optimal Control for Jump Diffusion Models with Singular Drifts
-
Low precision triggers slingshot loss spikes via feature inflation
Grokking or Glitching? How Low-Precision Drives Slingshot Loss Spikes
-
Floating-point limits trigger slingshot loss spikes
Grokking or Glitching? How Low-Precision Drives Slingshot Loss Spikes
-
Rates derived for stochastic quasi-Fejér sequences in metric spaces
Convergence guarantees for stochastic algorithms solving non-unique problems in metric spaces
-
Gradient feedback preserves NI property and stability
Absolute Stability of Nonlinear Negative Imaginary Systems with Application to Potential Energy Shaping
-
Optimized designs lift heat exchanger performance by 22 percent
Topology optimization of two-fluid turbulent heat exchangers: A Darcy flow-based multifidelity approach
-
Progressive IP finds local solutions for Heaviside composite problems
Solving Constrained Affine Heaviside Composite Optimization Problems by a Progressive IP Approach