en
0.25
0.5
0.75
1.25
1.5
1.75
2
Minimum Neighbor Distance Estimators of Intrinsic Dimension
Published on 2011-11-302562 Views
Most of the machine learning techniques suffer the "curse of dimensionality" effect when applied to high dimensional data. To face this limitation, a common preprocessing step consists in employing a
Related categories
Presentation
Minimum Neighbor Distance Estimators of Intrinsic Dimension00:00
Outline00:13
Motivation - 100:34
Motivation - 200:39
Motivation - 300:44
Motivation - 400:58
Motivation - 501:10
Motivation - 601:16
Motivation - 701:21
Applications - 101:29
Applications - 201:36
Applications - 301:41
Problems arising with dimensionality - 101:50
Problems arising with dimensionality - 202:07
Problems arising with dimensionality - 302:23
Problems arising with dimensionality - 402:40
Dimensionality estimation algorithms - 103:18
Dimensionality estimation algorithms - 203:37
Dimensionality estimation algorithms - 303:45
Some state of the art techniques - 104:05
Some state of the art techniques - 204:11
Some state of the art techniques - 304:20
Some state of the art techniques - 404:38
Some considerations - 104:43
Some considerations - 204:57
Some considerations - 305:00
Some considerations - 405:04
Some considerations - 505:10
Some considerations - 605:16
Some considerations - 705:24
Our approach - 105:37
Our approach - 205:50
Our approach - 305:56
Our approach - 406:08
Our approach - 506:20
Local uniformity - 106:36
Local uniformity - 206:55
A log-likelihood function - 107:06
A log-likelihood function - 207:17
A log-likelihood function - 307:26
A log-likelihood function - 407:39
Log-likelihood - 107:53
Log-likelihood - 207:57
Log-likelihood - 308:14
Log-likelihood - 408:27
pdf comparison - 108:39
pdf comparison - 208:47
pdf comparison - 308:58
pdf comparison - 409:04
pdf comparison - 509:12
MiNDKL - 109:19
MiNDKL - 209:31
MiNDKL - 309:34
MiNDKL - 409:42
Tests - 109:51
Tests - 209:55
Tests - 310:00
Experimental Setting - 110:13
Experimental Setting - 210:21
Results - 110:30
Results - 210:56
Conclusions - 111:41
Conclusions - 211:53
Conclusions - 311:55
Conclusions - 412:04
Future Works - 112:16
Future Works - 212:21
Questions12:32