sl
0.25
0.5
0.75
1.25
1.5
1.75
2
Prepoznavanje čustev na Twitterju
Published on Dec 05, 20171108 Views
Ste kdaj želeli prepoznavati čustva v tvitih, pa je teh bilo preveč za ročno pregledovanje? Bi vam koristilo, če bi računalnik sam znal ugotoviti, ali je bil nekdo jezen ali vesel, ko je nekaj zapisal
Related categories
Chapter list
Outline00:12
The problem01:05
The task01:47
The training data - 104:07
The training data - 205:52
Content filtering08:09
# depth filtering08:37
The training data - 308:53
Class distributions11:08
From text to features - 111:24
From text to features - 212:52
Results bow models13:18
From text to features16:27
DL architecture22:04
Results DL23:02
Transfer27:59
How to train unison? - 133:04
Common model - unison34:29
Alternate batches strategy34:42
How to train unison? - 235:07
Weighted sampling batches strategy35:40
Weighted sampling batches sampling probabilites35:58
Unison results36:55
T-SNE on char embeddings39:20
Take-away messages40:33