ICML 2023 accepted papers on Learning theory and algorithms
283 of the 1,865 papers accepted at ICML 2023 (15.2%) are on Learning theory and algorithms, compared with 15.6% at ICML 2022. They fall into 9 topics.
See ICML 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 · 83 papers
- Cross-Entropy Loss Functions: Theoretical Analysis and Applications (217 citations)
- Compute-Efficient Deep Learning: Algorithmic Trends and Opportunities (25 citations)
- Faith-Shap: The Faithful Shapley Interaction Index (19 citations)
- Theoretical Behavior of XAI Methods in the Presence of Suppressor Variables (8 citations)
- abess: A Fast Best-Subset Selection Library in Python and R (7 citations)
- Flexible Model Aggregation for Quantile Regression (7 citations)
- Constrained Monotonic Neural Networks (6 citations)
- Learning Optimal Group-structured Individualized Treatment Rules with Many Treatments (6 citations)
- ModelDiff: A Framework for Comparing Learning Algorithms (5 citations)
- The Numerical Stability of Hyperbolic Representation Learning (5 citations)
- Data-OOB: Out-of-bag Estimate as a Simple and Efficient Data Value (4 citations)
- Towards Bridging the Gaps between the Right to Explanation and the Right to be Forgotten (4 citations)
and 71 more in the interactive map.
Sample complexity · 51 papers
- Diffusion Models are Minimax Optimal Distribution Estimators (8 citations)
- Score Approximation, Estimation and Distribution Recovery of Diffusion Models on Low-Dimensional Data (8 citations)
- Are Gaussian Data All You Need? The Extents and Limits of Universality in High-Dimensional Generalized Linear Estimation (4 citations)
- Improved Analysis of Score-based Generative Modeling: User-Friendly Bounds under Minimal Smoothness Assumptions (4 citations)
- On Distribution Dependent Sub-Logarithmic Query Time of Learned Indexing (3 citations)
- Mixture Proportion Estimation Beyond Irreducibility (2 citations)
- A Law of Robustness beyond Isoperimetry (1 citation)
- AdaBoost is not an Optimal Weak to Strong Learner (1 citation)
- Effective Minkowski Dimension of Deep Nonparametric Regression: Function Approximation and Statistical Theories (1 citation)
- Learning Distributions over Quantum Measurement Outcomes (1 citation)
- Learning in POMDPs is Sample-Efficient with Hindsight Observability (1 citation)
- Multi-Fidelity Covariance Estimation in the Log-Euclidean Geometry (1 citation)
and 39 more in the interactive map.
Kernels and regression theory · 35 papers
- How Does Information Bottleneck Help Deep Learning? (14 citations)
- A Toy Model of Universality: Reverse Engineering how Networks Learn Group Operations (12 citations)
- Transformers as Algorithms: Generalization and Stability in In-context Learning (11 citations)
- Bayes-optimal Learning of Deep Random Networks of Extensive-width (6 citations)
- Deterministic equivalent and error universality of deep random features learning (4 citations)
- Deep linear networks can benignly overfit when shallow ones do (3 citations)
- Generalization on the Unseen, Logic Reasoning and Degree Curriculum (3 citations)
- Hypothesis Transfer Learning with Surrogate Classification Losses: Generalization Bounds through Algorithmic Stability (3 citations)
- PAC-Bayesian Generalization Bounds for Adversarial Generative Models (3 citations)
- Tighter Information-Theoretic Generalization Bounds from Supersamples (3 citations)
- Why does Throwing Away Data Improve Worst-Group Error? (3 citations)
- A Fast, Well-Founded Approximation to the Empirical Neural Tangent Kernel (2 citations)
and 23 more in the interactive map.
Combinatorial and search optimization · 24 papers
- Constrained Efficient Global Optimization of Expensive Black-box Functions (11 citations)
- Searching Large Neighborhoods for Integer Linear Programs with Contrastive Learning (8 citations)
- Multi-Objective GFlowNets (3 citations)
- SurCo: Learning Linear SURrogates for COmbinatorial Nonlinear Optimization Problems (3 citations)
- Alternating Local Enumeration (TnALE): Solving Tensor Network Structure Search with Fewer Evaluations (2 citations)
- Approximately Optimal Core Shapes for Tensor Decompositions (2 citations)
- LinSATNet: The Positive Linear Satisfiability Neural Networks (2 citations)
- Automated Search for Conjectures on Mathematical Constants using Analysis of Integer Sequences (1 citation)
- Controllable Neural Symbolic Regression (1 citation)
- Maximum Optimality Margin: A Unified Approach for Contextual Linear Programming and Inverse Linear Programming (1 citation)
- Multi-Objective Population Based Training (1 citation)
- New metrics and search algorithms for weighted causal DAGs (1 citation)
and 12 more in the interactive map.
Matrix and tensor methods · 21 papers
- A Deep Conjugate Direction Method for Iteratively Solving Linear Systems (2 citations)
- Random Matrix Analysis to Balance between Supervised and Unsupervised Learning under the Low Density Separation Assumption (2 citations)
- Sketching Meets Differential Privacy: Fast Algorithm for Dynamic Kronecker Projection Maintenance (2 citations)
- Tensor Gaussian Process with Contraction for Multi-Channel Imaging Analysis (2 citations)
- Block Subsampled Randomized Hadamard Transform for Nyström Approximation on Distributed Architectures (1 citation)
- Building Neural Networks on Matrix Manifolds: A Gyrovector Space Approach (1 citation)
- Regression with Label Permutation in Generalized Linear Model (1 citation)
- Under-Counted Tensor Completion with Neural Incorporation of Attributes (1 citation)
- A Nearly-Optimal Bound for Fast Regression with $\ell_\infty$ Guarantee
- Adaptive Estimation of Graphical Models under Total Positivity
- Extending Kernel PCA through Dualization: Sparsity, Robustness and Fast Algorithms
- Fast $(1+\varepsilon)$-Approximation Algorithms for Binary Matrix Factorization
and 9 more in the interactive map.
Fairness · 21 papers
- Fair and Optimal Classification via Post-Processing (6 citations)
- Fair and Robust Estimation of Heterogeneous Treatment Effects for Policy Learning (5 citations)
- FAIRER: Fairness as Decision Rationale Alignment (4 citations)
- FARE: Provably Fair Representation Learning with Practical Certificates (4 citations)
- When Personalization Harms Performance: Reconsidering the Use of Group Attributes in Prediction (3 citations)
- Omnipredictors for Constrained Optimization (2 citations)
- Weak Proxies are Sufficient and Preferable for Fairness with Missing Sensitive Attributes (2 citations)
- Fairness in Matching under Uncertainty (1 citation)
- Learning Antidote Data to Individual Unfairness (1 citation)
- Fair Neighbor Embedding
- How to address monotonicity for model risk management?
- Improving Fair Training under Correlation Shifts
and 9 more in the interactive map.
Clustering · 19 papers
- Approximation Algorithms for Fair Range Clustering (4 citations)
- High-dimensional Clustering onto Hamiltonian Cycle (3 citations)
- Cluster Explanation via Polyhedral Descriptions (2 citations)
- Generalized Reductions: Making any Hierarchical Clustering Fair and Balanced with Low Cost (2 citations)
- Nearly-Optimal Hierarchical Clustering for Well-Clustered Graphs (2 citations)
- Differentially Private Hierarchical Clustering with Provable Approximation Guarantees (1 citation)
- Likelihood Adjusted Semidefinite Programs for Clustering Heterogeneous Data (1 citation)
- Optimal LP Rounding and Linear-Time Approximation Algorithms for Clustering Edge-Colored Hypergraphs (1 citation)
- Total Variation Graph Neural Networks (1 citation)
- Fast Combinatorial Algorithms for Min Max Correlation Clustering
- Multi-class Graph Clustering via Approximated Effective $p$-Resistance
- Near-Optimal Quantum Coreset Construction Algorithms for Clustering
and 7 more in the interactive map.
Submodular and game algorithms · 19 papers
- Submodular Order Functions and Assortment Optimization (7 citations)
- Statistical Inference and A/B Testing for First-Price Pacing Equilibria (6 citations)
- Fully Dynamic Submodular Maximization over Matroids (5 citations)
- How Bad is Top-$K$ Recommendation under Competing Content Creators? (5 citations)
- Achieving Hierarchy-Free Approximation for Bilevel Programs with Equilibrium Constraints (2 citations)
- Addressing Budget Allocation and Revenue Allocation in Data Market Environments Using an Adaptive Sampling Algorithm (2 citations)
- Fairness in Streaming Submodular Maximization over a Matroid Constraint (2 citations)
- Quantum Speedups for Zero-Sum Games via Improved Dynamic Gibbs Sampling (2 citations)
- Difference of submodular minimization via DC programming (1 citation)
- Abstracting Imperfect Information Away from Two-Player Zero-Sum Games
- Are Equivariant Equilibrium Approximators Beneficial?
- Auxiliary Learning as an Asymmetric Bargaining Game
and 7 more in the interactive map.
Optimal transport · 10 papers
- The multimarginal optimal transport formulation of adversarial multiclass classification (7 citations)
- Minimax estimation of discontinuous optimal transport maps: The semi-discrete case (5 citations)
- The Monge Gap: A Regularizer to Learn All Transport Maps (4 citations)
- Monge, Bregman and Occam: Interpretable Optimal Transport in High-Dimensions with Feature-Sparse Maps (1 citation)
- Shedding a PAC-Bayesian Light on Adaptive Sliced-Wasserstein Distances (1 citation)
- Sliced-Wasserstein on Symmetric Positive Definite Matrices for M/EEG Signals (1 citation)
- Linear optimal partial transport embedding
- Meta Optimal Transport
- On Kinetic Optimal Probability Paths for Generative Models
- Recovery Bounds on Class-Based Optimal Transport: A Sum-of-Norms Regularization Framework