Augmented-SVM: Automatic space partitioning for combining multiple non-linear dynamics thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Augmented-SVM: Automatic space partitioning for combining multiple non-linear dynamics

Published on Jan 16, 20134329 Views

Non-linear dynamical systems (DS) have been used extensively for building generative models of human behavior. Its applications range from modeling brain dynamics to encoding motor commands. Many sche

Related categories

Chapter list

Augmented-SVM: Automatic space partitioning for combining multiple non-linear dynamics00:00
Dynamical Systems00:35
Single attractor dynamics02:33
Multiple attractor dynamics03:33
Combination of dynamics04:49
Boundary avoiding06:10
Designing the classifier function: Augmented-SVM07:30
A-SVM Dual08:52
SMO & -variant for A-SVM10:08
Classifier function11:26
1-vs-rest multi-class implementation13:29
Application : reach-to-grasp motions (1)13:51
Application : reach-to-grasp motions (2)14:24
Application : reach-to-grasp motions (3)14:43
Application : reach-to-grasp motions (4)14:52
Fast decision-making + trajectory generation15:23
W5315:56