[HN Gopher] Machine Learning Model Homotopy
___________________________________________________________________
Machine Learning Model Homotopy
Author : jmount
Score : 63 points
Date : 2024-09-17 21:29 UTC (5 days ago)
(HTM) web link (win-vector.com)
(TXT) w3m dump (win-vector.com)
| djoldman wrote:
| References:
|
| https://github.com/WinVector/Examples/blob/main/Model_Homoto...
| disgruntledphd2 wrote:
| Yeah the GitHub link is the meat of the article. Funny stuff. I
| also would have assumed that the estimates wouldn't vary as
| much, naively.
| orimirs wrote:
| Is the title wrong? No mentions of homologies in the link
| disgruntledphd2 wrote:
| Follow the GitHub link.
| HeatrayEnjoyer wrote:
| The title changed
| bbstats wrote:
| The grammar in this makes it pretty hard to follow
| esafak wrote:
| Looks like the start of an interesting article. Needs
| development.
| yldedly wrote:
| Do you still see such dynamics in the coefficients if you have an
| order of magnitude more data or fewer dimensions? 100 points in
| 6D is not much even for linear regression, the model might just
| be too high variance to interpolate monotonically between the two
| populations.
| jmount wrote:
| The amount of data rapidly factors-out. One cat get the effect
| even for millions of points in 6d. It is the complexity of the
| model solution which is going to be degree D-1 polynomials over
| a shared degree D-1 denominator that drive the effect.
| mnky9800n wrote:
| I may be misunderstanding, but wouldn't you see this effect
| any time you did a k fold cross validation?
___________________________________________________________________
(page generated 2024-09-22 23:00 UTC)