The anomaly score is based on all encode fields, not just the field being predicted. See https://www.youtube.com/watch?v=XK5Dd8fGO2w for a recent discussion about it.
Also, the raw_anomaly_score == "anomaly score" and anomaly_score == "anomaly likelihood". I know that is confusing. For details about the difference between the two, see: https://www.youtube.com/watch?v=nVCKjZWYavM . I hope that helps you out, --------- Matt Taylor OS Community Flag-Bearer Numenta On Sun, Oct 11, 2015 at 5:13 AM, chandrasekhar s <[email protected] > wrote: > Hello NUPIC, > > I'm new to platform and been trying out for last few weeks. I set up the > traffic tutorial learn the > basics and now get hands on, i'm trying create an app using the air now > data using the skeleton app. > > Please help me with some questions, May be the questions are already > answered in some links/wiki, if so > please point to that > > 1. In the traffic example, the anomaly detection is based on one field say > Speed. My question if in my > use case I need to consider two points say just as an example Ozone and > Particle Pollution together to > treat to take a decision whether its an anomaly, how can that be done or > is anomaly computed always for > one field > > 2. I was checking data in 'metric' table in MySQL. What is the difference > between raw_anomaly_score & > anomaly_score. > > Best Regards > Chandra > > <https://sigads.rediff.com/RealMedia/ads/click_nx.ads/www.rediffmail.com/signatureline.htm@Middle?> > Get your own *FREE* website, *FREE* domain & *FREE* mobile app with > Company email. > *Know More >* > <http://track.rediff.com/click?url=___http://businessemail.rediff.com?sc_cid=sign-1-10-13___&cmp=host&lnk=sign-1-10-13&nsrv1=host>
