I have EEG time data from two groups of people - survivors and deaths. I would like to make a rule for what differentiates the signals in these two groups. My tools are SAS Enterprise Guide and SAS Enterprise Miner. I thought to use the following approach:
1. Separate time series into two groups in Miner using various available approaches
2. Score how well those groups correspond to actual groups of survivors and deaths
The first point seems like a fishing expedition. Can anyone with more experience suggest a tactics?
Can the second point be done in Miner? How?
All suggestions are welcome. Thank you very much for your help.