Bayesian Optimization in a Billion Dimensions via Random Embeddings thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Bayesian Optimization in a Billion Dimensions via Random Embeddings

Published on Nov 07, 20136154 Views

Bayesian optimization techniques have been successfully applied to robotics, planning, sensor placement, recommendation, advertising, intelligent user interfaces and automatic algorithm configuration.

Related categories

Chapter list

Bayesian optimization in high dimensions00:00
Black-box optimization00:17
Bayesian optimization - 101:20
Bayesian optimization - 205:09
Parameter07:30
Some acquisition functions07:46
Convergence of Bayes Opt (e.g. dF, Zoghi & Smola, 2012)18:06
Automatic Information extraction18:16
Automatic (Adaptive) Monte Carlo samplers - 120:02
Automatic (Adaptive) Monte Carlo samplers - 221:51
Analytics, dynamic creative content and A/B testing23:38
Animation session24:25
Sensor networks - 125:57
Sensor networks - 226:09
Automatic machine learning27:04
Other applications27:49
Scaling to high-dimensions28:25
Random Embedding Bayesian Optimization28:32
Algorithm29:43
Low and High dimension kernels30:41
Existence of minimum in low dimension31:28
Size of constrained space Y31:39
Regret depends on the lowdimension31:44
Rotation invariance31:50
Scaling to over a billion dimensions32:32
Tuning NP hard problem solvers33:03
Software33:54