Multiclass Learning Approaches: A Theoretical Comparison with Implications 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

Multiclass Learning Approaches: A Theoretical Comparison with Implications

Published on Jan 14, 20132891 Views

We theoretically analyze and compare the following five popular multiclass classification methods: One vs. All, All Pairs, Tree-based classifiers, Error Correcting Output Codes (ECOC) with randomly ge

Related categories

Chapter list

Multiclass Learning Approaches: A Theoretical Comparison with Implications00:00
How to analyse?00:56
Summary of Results02:22
Proof Techniques03:10