I am looking for large data sets for regression analysis (not for
classification)- in fact I would like to compare
a new idea for decision trees with some well-known procedures.
I have several good data sets (millions of obs and thousand of
predictors), but all of them are similar (prediction
of next year health care cost) and I need more variety.
The web sites I checked out (UCI; KDD Cup etc) did not have what I need:
real/generated data for regression analysis with at least 5-10,000
observations (more better) and 20+ predictors (more better).
It would be great if some of you give me a link to download such data
sets or idea where I could find them.
Thank you in advance
Ognian Asparouhov, Ph.D.
AI Insight, Inc.