SOTAVerified

Learning Theory

Learning theory

Papers

Showing 376400 of 852 papers

TitleStatusHype
Optimal Rejection Function Meets Character Recognition Tasks0
Approximability and Generalisation0
Don't fear the unlabelled: safe semi-supervised learning via simple debiasingCode3
PACTran: PAC-Bayesian Metrics for Estimating the Transferability of Pretrained Models to Classification TasksCode1
Rate-Distortion Theoretic Generalization Bounds for Stochastic Learning Algorithms0
A Characterization of Multiclass Learnability0
Contrasting random and learned features in deep Bayesian linear regression0
Towards a unified view of unsupervised non-local methods for image denoising: the NL-Ridge approachCode0
Generalized Label Shift Correction via Minimum Uncertainty Principle: Theory and Algorithm0
Towards Unsupervised Domain Adaptation via Domain-Transformer0
Information-Theoretic Analysis of Minimax Excess Risk0
Artificial Intelligence and Spontaneous Collusion0
Minimax Regret Optimization for Robust Machine Learning under Distribution Shift0
Deep Learning in Random Neural Fields: Numerical Experiments via Neural Tangent KernelCode0
Trust in AI: Interpretability is not necessary or sufficient, while black-box interaction is necessary and sufficient0
Smoothed Online Learning is as Easy as Statistical Learning0
The no-free-lunch theorems of supervised learning0
Interplay between depth of neural networks and locality of target functions0
Limitation of Characterizing Implicit Regularization by Data-independent Functions0
Universal Online Learning with Unbounded Losses: Memory Is All You Need0
Learning deterministic hydrodynamic equations from stochastic active particle dynamics0
Deep Learning Optimization Theory - Trajectory Analysis of Gradient Descent0
Understanding Few-Shot Multi-Task Representation Learning Theory0
Universal Online Learning: an Optimistically Universal Learning Rule0
On generalization bounds for deep networks based on loss surface implicit regularizationCode0
Show:102550
← PrevPage 16 of 35Next →

No leaderboard results yet.