I have a history of a process represented as a string. Each symbol in a string represents a distinct state of a process in a given month. I have hundreds of thousands of these. Below are 2 examples:

CCCCCCCCCCCC33CCCC3CCCCCCMFBBBBM999999999999999999963CM3CCCCCCCCCC
BFFFFFFFFFFFFFFFFFFFFFFFFF99999999999963333CC33333333CCCCCCCCCCCCCCCCCCCC

I'm looking to use a machine learning algorithm that could learn form the data, recognize patterns, and would be able to predict probabilities of future status of a process based on its past performance. Is it possible? What kind of methods should I look into?

asked Jan 15 at 13:25

andr111's gravatar image

andr111
1112

edited Jan 15 at 13:33

Be the first one to answer this question!
toggle preview

powered by OSQA

User submitted content is under Creative Commons: Attribution - Share Alike; Other things copyright (C) 2010, MetaOptimize LLC.