So the COLT conference started this morning in sweltering Paris. Many of the presentations have been featured in a fashion or another on Nuit Blanche. Here are the full proceedings:
Regular Papers
- On Consistent Surrogate Risk Minimization and Property Elicitation[abs] [pdf]
- Online Learning with Feedback Graphs: Beyond Bandits [abs] [pdf]
- Learning Overcomplete Latent Variable Models through Tensor Methods [abs] [pdf]
- Simple, Efficient, and Neural Algorithms for Sparse Coding [abs] [pdf]
- Label optimal regret bounds for online local learning [abs] [pdf]
- Efficient Learning of Linear Separators under Bounded Noise, [abs] [pdf]
- Efficient Representations for Lifelong Learning and Autoencoding [abs] [pdf]
- Optimally Combining Classifiers Using Unlabeled Data[abs] [pdf]
- Minimax Fixed-Design Linear Regression [abs] [pdf]
- Escaping the Local Minima via Simulated Annealing: Optimization of Approximately Convex Functions [abs] [pdf]
-
Bandit Convex Optimization:
T√ Regret in One Dimension [abs] [pdf] - The entropic barrier: a simple and optimal universal self-concordant barrier [abs] [pdf]
- Optimum Statistical Estimation with Strategic Data Sources [abs] [pdf]
- On the Complexity of Learning with Kernels [abs] [pdf]
- Learnability of Solutions to Conjunctive Queries: The Full Dichotomy [abs] [pdf]
- Sequential Information Maximization: When is Greedy Near-optimal? [abs] [pdf]
- Efficient Sampling for Gaussian Graphical Models via Spectral Sparsification [abs] [pdf]
- Stochastic Block Model and Community Detection in Sparse Graphs: A spectral algorithm with optimal rate of recovery [abs] [pdf]
- On-Line Learning Algorithms for Path Experts with Non-Additive Losses [abs] [pdf]
- Truthful Linear Regression, [abs] [pdf]
- A PTAS for Agnostically Learning Halfspaces [abs] [pdf]
- S2: An Efficient Graph Based Active Learning Algorithm with Application to Nonparametric Classification [abs] [pdf]
- Improved Sum-of-Squares Lower Bounds for Hidden Clique and Hidden Submatrix Problems [abs] [pdf]
- Contextual Dueling Bandits [abs] [pdf]
- Beyond Hartigan Consistency: Merge Distortion Metric for Hierarchical Clustering [abs] [pdf]
- Faster Algorithms for Testing under Conditional Sampling [abs] [pdf]
- Learning and inference in the presence of corrupted inputs [abs] [pdf]
- From Averaging to Acceleration, There is Only a Step-size [abs] [pdf]
- Variable Selection is Hard [abs] [pdf]
- Vector-Valued Property Elicitation [abs] [pdf]
- Competing with the Empirical Risk Minimizer in a Single Pass [abs] [pdf]
- A Chaining Algorithm for Online Nonparametric Regression [abs] [pdf]
- Escaping From Saddle Points — Online Stochastic Gradient for Tensor Decomposition [abs] [pdf]
- Learning the dependence structure of rare events: a non-asymptotic study [abs] [pdf]
- Thompson Sampling for Learning Parameterized Markov Decision Processes [abs] [pdf]
- Computational Lower Bounds for Community Detection on Random Graphs [abs] [pdf]
- Adaptive Recovery of Signals by Convex Optimization [abs] [pdf]
- Tensor principal component analysis via sum-of-square proofs [abs] [pdf]
- Fast Exact Matrix Completion with Finite Samples [abs] [pdf]
- Exp-Concavity of Proper Composite Losses [abs] [pdf]
- On Learning Distributions from their Samples [abs] [pdf]
- MCMC Learning, [abs] [pdf]
- Online with Spectral Bounds, [abs] [pdf]
- Regret Lower Bound and Optimal Algorithm in Dueling Bandit Problem [abs] [pdf]
- Second-order Quantile Methods for Experts and Combinatorial Games [abs] [pdf]
- Hierarchical Label Queries with Data-Dependent Partitions [abs] [pdf]
- Algorithms for Lipschitz Learning on Graphs [abs] [pdf]
- Low Rank Matrix Completion with Exponential Family Noise [abs] [pdf]
- Bad Universal Priors and Notions of Optimality [abs] [pdf]
- Learning with Square Loss: Localization through Offset Rademacher Complexity [abs] [pdf]
- Achieving All with No Parameters: AdaNormalHedge [abs] [pdf]
- Lower and Upper Bounds on the Generalization of Stochastic Exponentially Concave Optimization [abs] [pdf]
- Correlation Clustering with Noisy Partial Information [abs] [pdf]
- Online Density Estimation of Bradley-Terry Models [abs] [pdf]
- First-order regret bounds for combinatorial semi-bandits [abs] [pdf]
- Norm-Based Capacity Control in Neural Networks [abs] [pdf]
- Cortical Learning via Prediction [abs] [pdf]
- Partitioning Well-Clustered Graphs: Spectral Clustering Works! [abs] [pdf]
- Batched Bandit Problems [abs] [pdf]
- Hierarchies of Relaxations for Online Prediction Problems with Evolving Constraints [abs] [pdf]
- Fast Mixing for Discrete Point Processes [abs] [pdf]
- Generalized Mixability via Entropic Duality [abs] [pdf]
- On the Complexity of Bandit Linear Optimization [abs] [pdf]
- An Almost Optimal PAC Algorithm [abs] [pdf]
- Minimax rates for memory-bounded sparse linear regression [abs] [pdf]
- Interactive Fingerprinting Codes and the Hardness of Preventing False Discovery [abs] [pdf]
- Convex Risk Minimization and Conditional Probability Estimation [abs] [pdf]
- Regularized Linear Regression: A Precise Analysis of the Estimation Error, [abs] [pdf]
- Max vs Min: Tensor Decomposition and ICA with nearly Linear Sample Complexity, [abs] [pdf]
- On Convergence of Emphatic Temporal-Difference Learning[abs] [pdf]
Open Problems
- Open Problem: Restricted Eigenvalue Condition for Heavy Tailed Designs, [abs] [pdf]
- Open Problem: The landscape of the loss surfaces of multilayer networks [abs] [pdf]
- Open Problem: The Oracle Complexity of Smooth Convex Optimization in Nonstandard Settings [abs] [pdf]
- Open Problem: Online Sabotaged Shortest Path, [abs] [pdf]
- Open Problem: Learning Quantum Circuits with Queries, [abs] [pdf]
- Open Problem: Recursive Teaching Dimension Versus VC Dimension, [abs] [pdf]
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.
No comments:
Post a Comment