# Dustin Tran

New York, NY

dustintran.com

Age: 23

Computer Science Ph.D. student at Columbia. Researcher in machine learning and Bayesian statistics.

 Mar 22 comment why gradient descent when we can solve linear regression analyticallyThis is not necessarily the bottleneck. To even use the normal equations, we typically make a non-singular assumption so that \$n > p\$ (here I'm using the notation that \$n\$ is number of data points and \$p\$ is number of features). This means the bottleneck is \$O(np^2)\$ to form \$X^\top X\$, not the \$O(p^3)\$ inversion. Mar 13 awarded Popular Question Feb 18 awarded Curious Jun 14 awarded Supporter Apr 20 awarded Student Apr 20 asked What data generating process can't be expressed as a probabilistic graphical model? Apr 17 awarded Constituent Apr 14 awarded Caucus Apr 2 awarded Popular Question Mar 28 awarded Yearling Mar 28 awarded Yearling Mar 5 accepted Maximal Unramified Extension of \$\mathbb{F}_p((t))\$ Jan 4 comment Running tests over samples from cross-validation on multiple datasets. To average or not to average?Is this a homework question? As I've said, they're both bad. If you're looking for the better of two evils, you can use the same reasoning applied above to try to deduce this. Jan 2 comment Running tests over samples from cross-validation on multiple datasets. To average or not to average?In order to see if the model is statistically significant in general, you test against a null hypothesis, i.e., "random guessing". This can be reformulated as a null model, and thus what you now do is model selection comparing how different the two models are. If you want to do an averaging using CV as your comparison, then what I described above is the best way to do so. You form a \$t\$-test on one statistic, the mean of the \$k\$ errors. This is slightly problematic as I mentioned above, but not as much as the other proposed \$t\$-tests. Dec 31 awarded Teacher Dec 30 answered Can I include covariate as random effect in glmer? Dec 30 answered Running tests over samples from cross-validation on multiple datasets. To average or not to average? Dec 30 awarded Autobiographer Dec 16 awarded Caucus Dec 4 awarded Student