my collaboration with China
Jun. 23rd, 2008 09:04 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
guslI love the Internet!
A month ago, an economics PhD student in China read my UAI paper on discovering cyclic causal models from non-experimental equilibrium data. Since then, he has downloaded my code, and run it on some real econometric data (with a ton of variables!).
It's really cool to look at the output, even if I don't know what the variables are.
With 10 variables, the distribution-equivalent set output by LiNG-D had 240 models. But only 2 of those are stable. This demonstrates my point (in the "Discussion" section) that stability can be a powerful selection criterion.
A month ago, an economics PhD student in China read my UAI paper on discovering cyclic causal models from non-experimental equilibrium data. Since then, he has downloaded my code, and run it on some real econometric data (with a ton of variables!).
It's really cool to look at the output, even if I don't know what the variables are.
With 10 variables, the distribution-equivalent set output by LiNG-D had 240 models. But only 2 of those are stable. This demonstrates my point (in the "Discussion" section) that stability can be a powerful selection criterion.