0.25
0.5
0.75
1.25
1.5
1.75
2
Mining Relational Model Trees
Published on Feb 25, 20076354 Views
Multi-Relational Data Mining (MRDM) refers to the process of discovering implicit, previously unknown and potentially useful information from data scattered in multiple tables of a relational database
Related categories
Chapter list
Mining Relational Model Trees00:01
Regression problem in classical data mining00:45
Regression trees and model trees01:46
Model trees: state of the art02:53
Model trees: state of the art 03:50
Two types of nodes05:29
What is passed down?07:11
An example of model tree09:17
Building a regression model stepwise: some tricks10:18
The global effect of regression nodes12:31
Advantages of the proposed tree structure14:32
Evaluating splitting and regression nodes15:33
Filtering useless splitting nodes18:23
Stopping criteria19:10
Related works … and problems21:33
Related works … and problems22:32
Related works … and problems23:32
Related works … and problems24:44
Computational complexity25:31
Empirical evaluation26:25
…Empirical evaluation on laboratory-sized data…27:17
…Empirical evaluation on laboratory-sized data34:15
… Empirical Evaluation on UCI data…35:04
… Empirical Evaluation on UCI data.37:18
SMOTI: open issues38:47
From classical to relational data mining40:09
Multi-relational representation41:16
Regression Problem in relational data mining41:58
How to work with (multi-)relational data?43:14
Strengths and Weaknesses of current multi-relational regression methods47:40
Global/local effect+ multi-relational model =Mr-SMOTI49:34
What is a regression selection graph?51:39
Relational splitting nodes53:04
Relational splitting nodes54:35
Relational splitting nodes55:45
Relational splitting nodes with look-ahead56:47
Relational regression nodes57:56
Relational model trees: an example58:20
How to choose the best relational node?01:00:18
Evaluating relational splitting node01:00:55
Evaluating relational regression node01:02:05
Stopping criteria01:02:24
Mr-SMOTI: some details01:02:57
Empirical evaluation on laboratory-sized data01:03:05
Empirical evaluation on laboratory-sized data01:03:24
Empirical evaluation on laboratory-sized data01:05:47
Empirical evaluation on laboratory-sized data01:06:30
Empirical evaluation on real data01:06:33
Improving efficency01:09:36