S2E37: A Summer of Love History of Quant, Part 3
Quantitude
00:00
The Unnerving Truth About ML
In the 60s, Gog was trying to improve upon the numerical methods of resolving these likelihood problems with the factor analysis. When you start laying out the model in this more general way, there's no reason at all why you can't start imposing certain restrictions on these parameters - one of which is fixing it to zero. You over identify the system and it becomes a testable hypothesis. And that brings us to his 1969 psychometric paper.
Transcript
Play full episode
