Re: Let's get back to earthquakes
Posted by marc / berkeley on June 19, 2006 at 11:36:33:

I am curious to the definition of the dataset required to make the prediction, or any prediction. Seven days and you get one hit, does that qualify as 6 misses? or misses until a hit, so if it hits on the last day, it's still 6 misses and one hit.

How would you define the dataset. If I predict 1 and get 4, then what.

I have to admit that maybe I am making this harder than it seems.

Perhaps you could "steer" me in the direction you are pointing, because without the right definition on the problem, things could get more complex than they need to be.

The simple solution is always the elegant solution.


Follow Ups:
     ● Re: Let's get back to earthquakes - Roger Hunter  12:31:50 - 6/19/2006  (38524)  (0)