Learning from Multiple Sources by Matching Their Distributions thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Learning from Multiple Sources by Matching Their Distributions

Published on Dec 20, 2008615 Views

My talk will address problems of transfer learning in which auxiliary data sources produce observations that are potentially helpful, but are not governed by the target distribution of the task at han