Sign inSign up
Fast DTT - A Near Linear Algorithm for Decomposing a Tensor into Factor Tensors thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Fast DTT - A Near Linear Algorithm for Decomposing a Tensor into Factor Tensors

Published on Oct 07, 20142161 Views

Xiaomin Fang

As tensors provide a natural representation for the higher-order relations, tensor factorization techniques such as Tucker decomposition and CANDECOMP/PARAFAC decomposition have been applied to many f

Research Sessions

Related categories

Data MiningKnowledge Extraction

Chapter list

Fast DTT – A Near Linear Algorithm for Decomposing a Tensor into Factor Tensors00:00
Introduction00:12
Introduction - Matrix and Tensor00:35
Motivation - Matrix Decomposition01:09
Motivation - Example 02:10
Motivation - Traditional Tensor Decomposition - 102:44
Motivation - Traditional Tensor Decomposition - 203:17
Motivation - DTT: Decompose Tensor into Factor Tensors - 104:05
Motivation - DTT: Decompose Tensor into Factor Tensors - 204:39
Model - Tensor Reconstruction - 105:35
Model - Tensor Reconstruction - 206:12
Model - Tensor Reconstruction - 306:36
Model - Tensor Reconstruction - 406:48
Model - Tensor Reconstruction - 507:04
Model - Limitations of DTT07:44
Model - Fast DTT - 107:57
Model - Fast DTT - 208:47
Model - Comparison between Fast DTT, TD and CP09:13
Experiments - Experimental Setup10:04
Experiments - Comparative Methods10:46
Experiments - Synthetic Datasets11:16
Experiment - Performance on Synthetic Datasets12:04
Experiments - Real Datasets12:34
Experiments - Performance on Real Datasets13:01
Experiments - Influence of the Number of Features13:23
Conclusion14:06
Thank you!15:04