Statistical Learning
Dropping Symmetry for Fast Symmetric Nonnegative Matrix Factorization
Zhihui Zhu, Xiao Li, Kai Liu, Qiuwei Li
NMF is equivalent to the classical K -means kernel clustering in [11]and it is inherently suitable for clustering nonlinearly separable data from a similarity matrix [10].
Virtual Class Enhanced Discriminative Embedding Learning
Binghui Chen, Weihong Deng, Haifeng Shen
Neural Information Processing Systems http://nips.cc/
Spectral Filtering for General Linear Dynamical Systems
Elad Hazan, Holden Lee, Karan Singh, Cyril Zhang, Yi Zhang
Neural Information Processing Systems http://nips.cc/
Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation
Jiaxuan You, Bowen Liu, Zhitao Ying, Vijay Pande, Jure Leskovec
Generating novel graph structures that optimize given objectives while obeying some given underlying rules is fundamental for chemistry, biology and social science research.
Accelerated Stochastic Matrix Inversion: General Theory and Speeding up BFGS Rules for Faster Second-Order Optimization
Robert Gower, Filip Hanzely, Peter Richtarik, Sebastian U. Stich
Neural Information Processing Systems http://nips.cc/
NEON2: Finding Local Minima via First-Order Oracles
We propose a reduction for non-convex optimization that can (1) turn an stationary-point finding algorithm into an local-minimum finding one, and (2) replace the Hessian-vector product computations with only gradient computations. It works both in the stochastic and the deterministic settings, without hurting the algorithm's performance. As applications, our reduction turns Natasha2 into a first-order method without hurting its theoretical performance. It also converts SGD, GD, SCSG, and SVRG into algorithms finding approximate local minima, outperforming some best known results.
Learning Invariances using the Marginal Likelihood
Mark van der Wilk, Matthias Bauer, ST John, James Hensman
By constraining our predictor to make similar predictions for inputs which are modified in ways that are irrelevant to the prediction (e.g.
Preference Based Adaptation for Learning Objectives
In many real-world learning tasks, it is hard to directly optimize the true performance measures, meanwhile choosing the right surrogate objectives is also difficult. Under this situation, it is desirable to incorporate an optimization of objective process into the learning loop based on weak modeling of the relationship between the true measure and the objective. In this work, we discuss the task of objective adaptation, in which the learner iteratively adapts the learning objective to the underlying true objective based on the preference feedback from an oracle. We show that when the objective can be linearly parameterized, this preference based learning problem can be solved by utilizing the dueling bandit model.
SLANG: Fast Structured Covariance Approximations for Bayesian Deep Learning with Natural Gradient
Aaron Mishkin, Frederik Kunstner, Didrik Nielsen, Mark Schmidt, Mohammad Emtiyaz Khan
To address this issue, we propose a new stochastic, low-rank, approximate natural-gradient (SLANG) method for variational inference in large, deep models. Our method estimates a "diagonal