Just added the t-test plots with description and a brief discussion of why the beta results were better. I excluded the db-scan because it seemed to produce to many cluster assignments and really did not add anything to the key point about why we needed to use a different feature set to train on.
Just added the t-test plots with description and a brief discussion of why the beta results were better. I excluded the db-scan because it seemed to produce to many cluster assignments and really did not add anything to the key point about why we needed to use a different feature set to train on.