decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books

Gear

Groklaw Gear

Click here to send an email to the editor of this weblog.


You won't find me on Facebook


Donate

Donate Paypal


No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster
Actually | 443 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
Actually
Authored by: Anonymous on Monday, December 31 2012 @ 04:59 PM EST
it's still a matter of debate as to how much digital is in your brain. But in
any case, as I have repeatedly pointed out, the brain is simply not fast enough
for a lot of tasks.

[ Reply to This | Parent | # ]

The problem..
Authored by: jesse on Tuesday, January 01 2013 @ 10:31 AM EST
The problem is in the analog nature itself.

If I remember right, there are two problems:

1) calibration - Each operational unit must be calibrated to function within its
circuit. Each circuit must then be calibrated to function within the larger
circuits involving other units. Setting one unit can alter the functioning of
the other units.

2) maintaining calibration - over time, the parts wear (in/out), thus changing
their calibration. To compensate for this requires additional circuit elements
to identify and feed back the error into the original circuit (to keep them
stabilized). These additional elements also require calibration...

And thus become a recursive problem...

Even during WW2, the fanciest bomb sight (the Norden one) had to be kept at a
proper temperature, or the targeting would be off due to the
expansion/contraction of the gears causing inaccurate runs.

So calibration can only be done "close enough". But the result isn't
stable over time.

Part of neural net inputs is nothing but such calibration feedback. The problem
the research has is identifying/separating out which inputs are for
calibration... and which are the data. After all they both look the same.

Most of the neural net research being done on a digital computer ignore the
calibration -- other than calling it feedback learning.

The organic neural nets have more inputs than just the interconnection network.
They also have food/oxygen/hormone/chemical/magnetic/... inputs. For the most
part these seem to be ignored or classed as "internal state", but none
of the formula for neural nets (at least the ones I've been able to find)
include them.

[ Reply to This | Parent | # ]

Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )