Collecting interesting data sets


#1

It seems as if there is no list of possibly interesting data sets, is there?
As it is out homework to try random data sets maybe we should collect some links?

For starters:




#2

#3

Thank you for the initiative!

I also started Houses and Titanic, but I will definitely try Red Wine Quality as well.

Anyway, I have just submitted my first submission on House Prices. I got an rmse of 0.14846. Do you think that’s good? How did you score on that one?

Also, I have another question, when I predicted the prices, they were very different from the prices shown in the sample submission file. For example, for 1461 on the test set, I get 124830.788, while the price in the submission file is 169277.0524984. Are those random values in the submission sample ?

Finally, why did m.predict stop at 3 decimals while the submission file stops at 7 decimal points?