Let's Make Block Coordinate Descent Go Fast: Faster Greedy Rules, Message-Passing, Active-Set Complexity, and Superlinear Convergence

By Julie Nutini, Issam Hadj Laradji, Mark W. Schmidt. 2017

50 citations recorded by the source corpus.

Subfields: Optimization & Training

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Optimization & Training

1. Nicolas Loizou
2. Simon Lacoste-Julien
3. Peter Richtárik
4. Issam Hadj Laradji
5. Sharan Vaswani
6. Mark W. Schmidt
7. Sebastian U. Stich
8. Francis R. Bach

Citations in the corpus

• Stochastic Polyak Step-size for SGD: An Adaptive Learning Rate for Fast Convergence (2020, methodology): “Following the procedure outlined in Nutini et al. (2017), we generate a sparse dataset for binary classification with the number of examples n = 1k and dimension d = 100.”
• Coordinate Descent Converges Faster with the Gauss-Southwell Rule Than Random Selection (2015, methodology): “Under a suitable non-degeneracy assumption, coordinate descent methods achieve a particular “active set” property in a finite number of iterations [Wright, 2012, Nutini et al., 2017].”
• Where are the Blobs: Counting by Localization with Point Supervision (2018, methodology): “For example, fast label propagation methods proposed in [34,35] can be used to determine the boundaries between the objects in the image.”