Sign inSign up
Multi-task Regularization of Generative Similarity Models 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

Multi-task Regularization of Generative Similarity Models

Published on Oct 17, 20113006 Views

Luca Cazzanti

We investigate a multi-task approach to similarity discriminant analysis, where we propose treating the estimation of the different class-conditional distributions of the pairwise similarities as mult

SIMBAD 2011 - Venice

Related categories

Multi-Task Learning

Chapter list

Multi-­‐task Regularization of Generative Similarity Models00:00
Outline00:09
Euclidean Features00:57
Similarities - 101:29
Similarities - 202:00
Local Similarity Discriminant Analysis (local SDA) - 102:15
Local Similarity Discriminant Analysis (local SDA) - 202:43
Local Similarity Discriminant Analysis (local SDA) - 302:58
Local Similarity Discriminant Analysis (local SDA) - 403:10
Local Similarity Discriminant Analysis (local SDA) - 503:24
Local Similarity Discriminant Analysis (local SDA) - 603:31
Estimating the Local SDA Parameters - 104:07
Estimating the Local SDA Parameters - 204:43
Estimating the Local SDA Parameters - 304:49
Estimating the Local SDA Parameters - 404:55
Need for Regularization - 105:10
Need for Regularization - 205:52
Need for Regularization - 306:23
Multi-task Regularization - 106:49
Multi-task Regularization - 207:38
Multi-task Regularization – Closed Form Solution08:26
Choice of Task Relatedness Matrix A09:20
Benchmark Datasets - 110:19
Benchmark Datasets - 211:55
Benchmark Datasets - 312:41
Insurgent Rhetoric Experiment13:15
Summary15:43