Quantitude

S4E10 Test Driving Model Identification

Nov 29, 2022
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 2min
2
Do You Have the Earring Yet?
01:40 • 2min
3
Model Identification - You Have to Have Enough, Enough Something
03:53 • 2min
4
The Six Shun Steps
05:23 • 2min
5
The Bottom Line Is a Model Can't Run if It's Not Identified
07:01 • 2min
6
The Problem of Identification in the Modeling World
08:32 • 2min
7
The Ferrari Implications Are Self Evident
10:58 • 3min
8
Is There a Constraint to Zero in Modeling?
13:35 • 2min
9
Is There a Constraint That You Can't Fix to a Numerical Value?
15:50 • 2min
10
Quantitative Methods Is Really Pretty Easy Because Notice We Have Not Left Middle School Math
17:44 • 2min
11
How Does the Covariance Between X and Y Affect Variance?
19:14 • 4min
12
The Trickiness of Identification
22:55 • 3min
13
What Is a Condition That Is Not Sufficient?
25:57 • 3min
14
Using Covariance Algorithms, You're a Coward.
28:55 • 4min
15
Is It Okay to Introduce Latent Variables?
33:13 • 2min
16
What Are Latent Variables?
35:20 • 2min
17
How to Scale Your Latent Variables
37:48 • 3min
18
How Do You Establish the Identification of a Model?
40:18 • 4min
19
The Only Thing Worse Than Having No Map Is Having a Bad Map
44:16 • 3min
20
What's the Biggest Problem in Model Identification?
47:31 • 3min