January 10, 2007
Economists get a lot of credit for teasing out meaningful correlations in big buckets of data. But if they are simply running regressions on data sets, isn’t this something a computer could do automatically via data mining? With a giant dataset from the World Health Organization, I wonder if data mining algorithms can reveal some interesting correlations automatically? A quick Google Scholarsearch didn’t turn up any papers, but I’m sure it’s been done. It’s too obvious for noone to have tried. Anyway, this will be a chance to tinker with DB design, which I’ve never had to do before.