ICML 2017 accepted papers on Learning theory and algorithms
117 of the 435 papers accepted at ICML 2017 (26.9%) are on Learning theory and algorithms, compared with 30.7% at ICML 2016. They fall into 9 topics.
See ICML 2017 in the interactive map
Submitting to ICML? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Matrix and tensor methods · 33 papers
- No Spurious Local Minima in Nonconvex Low Rank Problems: A Unified Geometric Analysis (215 citations)
- Practical Gauss-Newton Optimisation for Deep Learning (104 citations)
- Geometry of Neural Network Loss Surfaces via Random Matrix Theory (90 citations)
- Algebraic Variety Models for High-Rank Matrix Completion (32 citations)
- Communication-efficient Algorithms for Distributed Stochastic Principal Component Analysis (27 citations)
- Nearly Optimal Robust Matrix Completion (19 citations)
- Theoretical Properties for Neural Networks with Weight Matrices of Low Displacement Rank (19 citations)
- On the Iteration Complexity of Support Recovery via Hard Thresholding Pursuit (18 citations)
- Doubly Accelerated Methods for Faster CCA and Generalized Eigendecomposition (15 citations)
- Frame-based Data Factorizations (13 citations)
- On Approximation Guarantees for Greedy Low Rank Optimization (13 citations)
- Toward Efficient and Accurate Covariance Matrix Estimation on Compressed Data (13 citations)
and 21 more in the interactive map.
Prediction and decision losses · 21 papers
- Axiomatic Attribution for Deep Networks (2,911 citations)
- Understanding Black-box Predictions via Influence Functions (1,328 citations)
- Learning Algorithms for Active Learning (107 citations)
- Gradient Boosted Decision Trees for High Dimensional Sparse Output (103 citations)
- Iterative Machine Teaching (92 citations)
- Joint Dimensionality Reduction and Metric Learning: A Geometric Take (47 citations)
- Active Learning for Cost-Sensitive Classification (42 citations)
- Optimal Densification for Fast and Accurate Minwise Hashing (35 citations)
- Just Sort It! A Simple and Effective Approach to Active Preference Learning (30 citations)
- Logarithmic Time One-Against-Some (27 citations)
- Forest-type Regression with General Losses and Robust Forest (20 citations)
- Consistency Analysis for Binary Classification Revisited (12 citations)
and 9 more in the interactive map.
Sample complexity · 21 papers
- Recovery Guarantees for One-hidden-layer Neural Networks (166 citations)
- Distributed Mean Estimation with Limited Communication (145 citations)
- Depth-Width Tradeoffs in Approximating Natural Functions with Neural Networks (98 citations)
- Being Robust (in High Dimensions) Can Be Practical (68 citations)
- Analytical Guarantees on Numerical Precision of Deep Neural Networks (62 citations)
- Algorithmic Stability and Hypothesis Complexity (48 citations)
- Uniform Convergence Rates for Kernel Density Estimation (45 citations)
- Maximum Selection and Ranking under Noisy Comparisons (22 citations)
- Neural Networks and Rational Functions (21 citations)
- Collect at Once, Use Effectively: Making Non-interactive Locally Private Learning Possible (19 citations)
- Tight Bounds for Approximate Carathéodory and Beyond (14 citations)
- Density Level Set Estimation on Manifolds with DBSCAN (13 citations)
and 9 more in the interactive map.
Clustering · 12 papers
- Deep Spectral Clustering Learning (104 citations)
- Multilevel Clustering via Wasserstein Means (85 citations)
- Clustering by Sum of Norms: Stochastic Incremental Algorithm, Convergence and Cluster Recovery (29 citations)
- Distributed and Provably Good Seedings for k-Means in Constant Rounds (19 citations)
- Clustering High Dimensional Dynamic Data Streams (17 citations)
- Consistent k-Clustering (14 citations)
- Composing Tree Graphical Models with Persistent Homology Features for Clustering Mixed-Type Data (12 citations)
- Innovation Pursuit: A New Approach to the Subspace Clustering Problem (11 citations)
- Multiple Clustering Views from Multiple Uncertain Experts (11 citations)
- Nyström Method with Kernel K-means++ Samples as Landmarks (11 citations)
- Uniform Deviation Bounds for k-Means Clustering (10 citations)
- Leveraging Union of Subspace Structure to Improve Constrained Clustering
Submodular and game algorithms · 9 papers
- Guarantees for Greedy Maximization of Non-submodular Functions with Applications (100 citations)
- Robust Submodular Maximization: A Non-Uniform Partitioning Approach (35 citations)
- Probabilistic Submodular Maximization in Sub-Linear Time (21 citations)
- Reduced Space and Faster Convergence in Imperfect-Information Games via Pruning (19 citations)
- Regret Minimization in Behaviorally-Constrained Zero-Sum Games (14 citations)
- Robust Guarantees of Stochastic Greedy Algorithms (9 citations)
- Adapting Kernel Representations Online Using Submodular Maximization (4 citations)
- Deletion-Robust Submodular Maximization: Data Summarization with “the Right to be Forgotten”
- Robust Budget Allocation via Continuous Submodular Functions
Combinatorial and search optimization · 9 papers
- Count-Based Exploration with Neural Density Models (343 citations)
- On Relaxing Determinism in Arithmetic Circuits (35 citations)
- Distributed Batch Gaussian Process Optimization (32 citations)
- Re-revisiting Learning on Hypergraphs: Confidence Interval and Subgradient Method (26 citations)
- Fast k-Nearest Neighbour Search via Prioritized DCI (19 citations)
- Strong NP-Hardness for Sparse Optimization with Concave Penalty Functions (13 citations)
- Rule-Enhanced Penalized Regression by Column Generation using Rectangular Maximum Agreement (2 citations)
- Exact MAP Inference by Avoiding Fractional Vertices (1 citation)
- On The Projection Operator to A Three-view Cardinality Constrained Set
Kernels and regression theory · 8 papers
- Random Fourier Features for Kernel Ridge Regression: Approximation Bounds and Statistical Guarantees (82 citations)
- Sketched Ridge Regression: Optimization Perspective, Statistical Perspective, and Model Averaging (50 citations)
- Kernelized Support Tensor Machines (40 citations)
- Learning Latent Space Models with Angular Constraints (17 citations)
- An Adaptive Test of Independence with Analytic Kernel Embeddings (16 citations)
- Spherical Structured Feature Maps for Kernel Approximation (11 citations)
- Learning Infinite Layer Networks Without the Kernel Trick (3 citations)
- Differentially Private Ordinary Least Squares
Optimal transport · 3 papers
- Sliced Wasserstein Kernel for Persistence Diagrams (151 citations)
- Co-clustering through Optimal Transport (7 citations)
- A Simulated Annealing Based Inexact Oracle for Wasserstein Loss Minimization (5 citations)
Fairness · 1 paper
- Meritocratic Fairness for Cross-Population Selection (37 citations)
