So the granularity of the Anomaly detection is across all inputs? Or am I understanding this wrongly? If so, this would mean that the anomaly detected just states that "something" was anomalous, but with no indication as to which attribute?
If all the above is correct, then is there a way to measure variation of *one* attribute of an aggregate input? If the above is *not* correct, then how does the anomaly calculation relate to each field of an compacted SDR, and how does one pull that information out? Or, are these concerns not addressed at this time? On Thu, Feb 12, 2015 at 1:06 PM, Scott Purdy <[email protected]> wrote: > No, the anomaly score is per-model. It is computed based on the temporal > memory state. > On Feb 12, 2015 10:58 AM, "cogmission" <[email protected]> wrote: > >> Scott, >> >> I'm resending this for clarity... >> >> Thanks for the prompt response. >> How about Anomaly instance usage? I would assume a single instance of an >> Anomaly tracks the anomaly scoring for a single unique field, correct? So >> then, there must be one per field entry type no? >> >> David >> > -- *We find it hard to hear what another is saying because of how loudly "who one is", speaks...*
