May 8, 2012
I would like to hear about large scale data mining without using traditional databases and full blown general purpose tools.
I am currently researching image similarity techniques, adding to a large body of work in time series data mining. I once generated and stored every reachable configuration of the Rubiks cube in under an hour on a single CPU core, with the resulting data structure allowing O(1) verification that a state is reachable.
No introduction yet