[Sigia-l] "Gait code"

Listera listera at rcn.com
Mon Oct 17 21:03:32 EDT 2005


Trenouth, John:

> Ziya said:
>> A lot of the pattern recognition/context analysis software
> 
> I'm having trouble believing that anyone would equate pattern recognition with
> context analysis.

Me, too. So why don't we get off the straw man argument or attribute to me
assertions I haven't made.

> What does the breathing/voice/speech/temperature mean?

It's data which when *combined* with, say, (tiny) facial muscle movements
(more data) and personal medical history (more data) and prior occurrence
(more data) and a whole bunch of other patterns that a machine can recognize
or have access to (even more data) can generate better/more cost-effective
results than a human can. These data/patterns are obviously weighed and
analyzed via some sort of algorithmic/rule-based system without which you'd
be looking at a fairly inert pile of data.

This isn't far from contextual analysis humans engage in: collect data, try
to detect patterns, weigh options/outcomes/etc and reach a conclusion. It's
just that machines can do this at a speed, depth and volume humans can't
match.

Are there areas where humans can do much, much better? Of course. Does that,
therefore, mean machines cannot do contextual analysis? Of course not. Have
we reached the apex of machine-based contextual analysis? Not by a few a
decades.  

---- 
Ziya

Best Practices,
For when you've run out of your own ideas and context.




More information about the Sigia-l mailing list