Sign inSign up
Beyond the regret minimization barrier: an optimal algorithm for stochastic strongly-convex optimization thumbnail
slide-image
Pause
Mute
Subtitles not available
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Beyond the regret minimization barrier: an optimal algorithm for stochastic strongly-convex optimization

Published on Aug 02, 20113626 Views

Satyen Kale

We give a novel algorithm for stochastic strongly-convex optimization in the gradient oracle model which returns an O(1/T)-approximate solution after T gradient updates. This rate of convergence is

COLT 2011 - Budapest

Related categories

Stochastic Optimization

Chapter list

Beyond the Regret Minimization Barrier00:00
Support Vector Machines00:00
Soft-Margin SVM - 100:37
Soft-Margin SVM - 200:38
Solving the SVM problem02:08
Stochastic Gradient Descent03:03
Optimization via Regret Minimization04:07
Stochastic Gradient Descent05:13
Natural Questions - 106:51
Natural Questions - 208:01
Epoch-GD08:47
Analysis Sketch - 109:41
Analysis Sketch - 211:26
Analysis Sketch - 311:59
Lower Bound Intuition12:51
Formalizing intuition14:36
Conclusions16:35
Thank you17:31