0.25
0.5
0.75
1.25
1.5
1.75
2
Modeling reality without sacrificing data: Inferentially tractable models for complex social systems
Published on Sep 26, 20084485 Views
Related categories
Chapter list
Modeling Reality Without Sacrificing Data: Inferentially Tractable Models for Complex Social Systems00:00
The Problem of Complex Dependence00:37
Challenge: Modeling Reality without Sacrificing Data01:40
Aside: Why “Small” Cases Can’t Be Ignored05:14
Today’s Case: Stochastic Models for Social (and Other) Networks06:50
Basic Notation07:46
Classical Examples09:05
Simple Relaxations11:51
A Less Trivial Model12:57
Exponential Families for Random Graphs14:52
Deduction with ERGs17:27
Example: Edge Clustering18:31
Simulating ERGs20:01
From Deduction to Inference21:54
Inference for ERGs23:40
Inference for ERGs, Cont. part125:44
Inference for ERGs, Cont. part226:31
Parameterizing ERGs27:44
Dependence Graphs and ERGs28:29
Model Construction Using Dependence Graphs30:59
Model Construction Using Dependence Graphs, Cont.31:48
A More Complex Example: The Markov Graphs32:20
Beyond the Markov Graphs34:03
Reciprocal Path Conditions35:27
Reciprocal Path Dependence Models36:44
Illustrative Application to Sample Networks37:34
Cycle Census ERG Fits38:19
A New Direction: Potential Games40:48
Potential Games and Network Formation Games42:01
Various Utility/Potential Components44:08
Empirical Example: Advice-Seeking Among Managers44:52
Advice-Seeking ERG – Model Comparison45:35
Advice-Seeking ERG – AIC Selected Model46:38
Model Adequacy Check48:30
Where Would One Go Next?49:32
Summary50:11
In Closing50:53