COSNet: a Cost Sensitive Neural Network for Semi-supervised Learning in Graphs thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

COSNet: a Cost Sensitive Neural Network for Semi-supervised Learning in Graphs

Published on Nov 30, 20112949 Views

The semi-supervised problem of learning node labels in graphs consists, given a partial graph labeling, in inferring the unknown labels of the unlabeled vertices. Several machine learning algorithms h

Related categories

Chapter list

COSNet: a Cost Sensitive Neural Network for Semi-supervised Learning in Graphs00:00
Outline00:48
Gene Function Prediction01:13
Gene Function Prediction Problem - 102:14
Gene Function Prediction Problem - 202:26
ecmlpkdd2011_frasca_cosnet_01_Page_0602:37
ecmlpkdd2011_frasca_cosnet_01_Page_0702:46
ecmlpkdd2011_frasca_cosnet_01_Page_0803:06
Machine learning methods for GFP03:09
Gene Annotation using Integrated Networks (GAIN)04:03
GAIN - 104:57
GAIN - 205:32
Drawbacks of GAIN - 106:27
Drawbacks of GAIN - 207:08
Parametrized Hopfield network - 107:49
Parametrized Hopfield network - 208:29
Sub Network09:01
Sub-network property - 110:10
Sub-network property - 210:57
Sketch of COSNet11:20
Generating a temporary solution12:44
Finding the optimal parameters - 113:09
Finding the optimal parameters - 214:32
Data imbalance management15:59
Finding the final solution17:24
Results - 118:02
Results -218:37
Results -319:39
Conclusions20:27
Ongoing developments21:43