Follow the Leader with Dropout Perturbations

By Tim van Erven, W. Kotłowski. 2014

55 citations recorded by the source corpus.

How researchers see this paper

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

1. Tal Ridnik
2. Asaf Noy
3. Niv Nayman
4. Lihi Zelnik-Manor
5. Itamar Friedman
6. Rong Jin
7. Hao Ma
8. Stanislav Fort

Citations in the corpus

• XNAS: Neural Architecture Search with Expert Advice (2019, background): “Common stochastic gradient-descent (SGD) and a set of PEA approaches, such as follow-theregularized-leader (FTRL), were shown by [36, 14, 43] to be equivalent.”
• On the inductive bias of dropout (2014, background): “…criterion has a unique minimizer, • when the dropout-regularization penalty goes to infinity with the weights, and when it remains bounded, • that the dropout regularization can be non-monotonic as individual weights increase from 0, and • that the dropout regularization penalty may not be convex.”
• Online Learning via the Differential Privacy Lens (2017, unknown): “Previous FTPL analysis achieving first-order regret bounds all relied on specific choices such as exponential [22] and dropout [41] perturbations.”