ICLR 2023 accepted papers on Learning theory and algorithms
123 of the 1,575 papers accepted at ICLR 2023 (7.8%) are on Learning theory and algorithms, compared with 8.5% at ICLR 2022. They fall into 9 topics.
See ICLR 2023 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Prediction and decision losses · 32 papers
- Probabilistically Robust Recourse: Navigating the Trade-offs between Costs and Robustness in Algorithmic Recourse (13 citations)
- Perfectly Secure Steganography Using Minimum Entropy Coupling (12 citations)
- Data Valuation Without Training of a Model (6 citations)
- LAVA: Data Valuation without Pre-Specified Learning Algorithms (6 citations)
- Understanding Why Generalized Reweighting Does Not Improve Over ERM (6 citations)
- HyperDeepONet: learning operator with complex target function space using the limited resources via hypernetwork (3 citations)
- On the Trade-Off between Actionable Explanations and the Right to be Forgotten (3 citations)
- Robust Explanation Constraints for Neural Networks (3 citations)
- SAM as an Optimal Relaxation of Bayes (3 citations)
- Efficiently Controlling Multiple Risks with Pareto Testing (2 citations)
- Interpretability with full complexity by constraining feature information (2 citations)
- Neural Bregman Divergences for Distance Learning (2 citations)
and 20 more in the interactive map.
Kernels and regression theory · 23 papers
- What learning algorithm is in-context learning? Investigations with linear models (86 citations)
- StableDR: Stabilized Doubly Robust Learning for Recommendation on Data Missing Not at Random (12 citations)
- Symmetric Pruning in Quantum Neural Networks (8 citations)
- The Onset of Variance-Limited Behavior for Networks in the Lazy and Rich Regimes (4 citations)
- Deep Learning meets Nonparametric Regression: Are Weight-Decayed DNNs Locally Adaptive? (3 citations)
- Generalization and Estimation Error Bounds for Model-based Neural Networks (3 citations)
- Theoretical Characterization of the Generalization Performance of Overfitted Meta-Learning (3 citations)
- Characterizing the spectrum of the NTK via a power series expansion (2 citations)
- Learning ReLU networks to high uniform accuracy is intractable (2 citations)
- On the Saturation Effect of Kernel Ridge Regression (2 citations)
- Contrastive Learning Can Find An Optimal Basis For Approximately View-Invariant Functions (1 citation)
- Minimax Optimal Kernel Operator Learning via Multilevel Training (1 citation)
and 11 more in the interactive map.
Combinatorial and search optimization · 15 papers
- Learning Cut Selection for Mixed-Integer Linear Programming via Hierarchical Sequence Model (14 citations)
- A GNN-Guided Predict-and-Search Framework for Mixed-Integer Linear Programming (10 citations)
- Solving Constrained Variational Inequalities via a First-order Interior Point-based Method (2 citations)
- Unsupervised Learning for Combinatorial Optimization Needs Meta Learning (2 citations)
- Dual Algorithmic Reasoning (1 citation)
- A Non-monotonic Self-terminating Language Model
- Approximate Vanishing Ideal Computations at Scale
- Multi-objective optimization via equivariant deep hypervolume approximation
- Ensuring DNN Solution Feasibility for Optimization Problems with Linear Constraints
- Generalize Learned Heuristics to Solve Large-scale Vehicle Routing Problems in Real-time
- Learning Sparse Group Models Through Boolean Relaxation
- Learning to CROSS exchange to solve min-max vehicle routing problems
and 3 more in the interactive map.
Optimal transport · 13 papers
- Neural Optimal Transport (5 citations)
- Sparsity-Constrained Optimal Transport (3 citations)
- Hierarchical Sliced Wasserstein Distance (2 citations)
- Riemannian Metric Learning via Optimal Transport (2 citations)
- On amortizing convex conjugates for optimal transport (1 citation)
- Spherical Sliced-Wasserstein (1 citation)
- A Convergent Single-Loop Algorithm for Relaxation of Gromov-Wasserstein in Graph Data
- Kernel Neural Optimal Transport
- Momentum Stiefel Optimizer, with Applications to Suitably-Orthogonal Attention, and Optimal Transport
- Ollivier-Ricci Curvature for Hypergraphs: A Unified Framework
- Computing all Optimal Partial Transports
- Efficient Discrete Multi Marginal Optimal Transport Regularization
and 1 more in the interactive map.
Sample complexity · 12 papers
- Nonlinear Reconstruction for Operator Learning of PDEs with Discontinuities (3 citations)
- Phase transition for detecting a small community in a large network (2 citations)
- On Achieving Optimal Adversarial Test Error (1 citation)
- Accelerating Hamiltonian Monte Carlo via Chebyshev Integration Time
- Almost Linear Constant-Factor Sketching for $\ell_1$ and Logistic Regression
- Depth Separation with Multilayer Mean-Field Networks
- On the complexity of nonsmooth automatic differentiation
- On the Performance of Temporal Difference Learning With Neural Networks
- Pitfalls of Gaussians as a noise distribution in NCE
- Provably Auditing Ordinary Least Squares in Low Dimensions
- Robust Algorithms on Adaptive Inputs from Bounded Adversaries
- A Higher Precision Algorithm for Computing the $1$-Wasserstein Distance
Fairness · 8 papers
- MEDFAIR: Benchmarking Fairness for Medical Imaging (14 citations)
- FIFA: Making Fairness More Generalizable in Classifiers Trained on Imbalanced Data (6 citations)
- Re-weighting Based Group Fairness Regularization via Classwise Robust Optimization (5 citations)
- Fairness and Accuracy under Domain Generalization (3 citations)
- Equal Improvability: A New Fairness Notion Considering the Long-term Impact (1 citation)
- FaiREE: fair classification with finite-sample and distribution-free guarantee
- Confidential-PROFITT: Confidential PROof of FaIr Training of Trees
- Measure the Predictive Heterogeneity
Submodular and game algorithms · 7 papers
- Modeling content creator incentives on algorithm-curated platforms (7 citations)
- Meta-Learning in Games (3 citations)
- $O(T^{-1})$ Convergence of Optimistic-Follow-the-Regularized-Leader in Two-Player Zero-Sum Markov Games (1 citation)
- ESCHER: Eschewing Importance Sampling in Games by Computing a History Value Function to Estimate Regret (1 citation)
- Statistical Inference for Fisher Market Equilibrium
- The Symmetric Generalized Eigenvalue Problem as a Nash Equilibrium
- Representation Learning for Low-rank General-sum Markov Games
Matrix and tensor methods · 7 papers
- Constructive TT-representation of the tensors given as index interaction functions with applications (3 citations)
- Learning Sparse and Low-Rank Priors for Image Recovery via Iterative Reweighted Least Squares Minimization (3 citations)
- Implicit Bias of Large Depth Networks: a Notion of Rank for Nonlinear Functions (2 citations)
- Learning the Positions in CountSketch
- Subquadratic Algorithms for Kernel Matrices via Kernel Density Estimation
- Tensor-Based Sketching Method for the Low-Rank Approximation of Data Streams.
- A Holistic View of Label Noise Transition Matrix in Deep Learning and Beyond
Clustering · 6 papers
- Improved Learning-augmented Algorithms for k-means and k-medians Clustering (3 citations)
- KwikBucks: Correlation Clustering with Cheap-Weak and Expensive-Strong Signals (3 citations)
- Learning Group Importance using the Differentiable Hypergeometric Distribution (1 citation)
- Near-optimal Coresets for Robust Clustering (1 citation)
- Mini-batch $k$-means terminates within $O(d/\epsilon)$ iterations
- Statistical Guarantees for Consensus Clustering