Re: What did that thread indicate?

On 29 Sep 2005 19:21:27 GMT, curt@xxxxxxxx (Curt Welch) wrote:

>Traveler <traveler@xxxxxxxxxx> wrote:
>> On 29 Sep 2005 05:07:17 GMT, curt@xxxxxxxx (Curt Welch) wrote:

[cut]

>If you have a audio stream of cocktail party noise, a general AI network
>needs the power to separate the single data stream into separate
>conversation. This is signal separation. It's a filtering processes. But
>when people say "filter" they normally mean, separate the signal into two
>parts, and then throw one part away. The general solution however needs
>"tunable filters" which spit signals.
>
>Taking a single radio signal and filtering into separate stations, is a
>signal separation processes. It can and must be done single signals and
>does not require two signals to do.
>
>Combining the audio streams togther in an audio mixer is an example of
>fusion. It's the exact inverse signal separation. It takes two signals,
>and combines them together.

What's with the analog signal crap? All analog signals are converted
to discrete signals in the brain by the sensor layer. By the time
signals arrive at the visual cortex, it's all discrete pulses. And all
along I thought we were talking about discrete signals. I even used
the term 'discrete signal separation' several times in my posts.

>> Likewise, signal fusion is not about merging two signals into a single
>> stream. Fusion is about finding coincident (simultaneous) signals and
>> combining them into a single signal. Both operations require a
>> temporal learning rule and a correlation factor. Signal fusion
>> necessarily must come after separation. That's because the correlation
>> factor would not work otherwise.
>
[cut]

>Everything you keep saying must be part of the system, I see happening in
>my net, and I see it happening in ways that seem far simpler and far more
>fundimential than what you seem to be doing (though you never give us full
>details of your algorthms so I never know for sure what you are in fact
>doing). You however, only seem to see your way of doing signal separation,
>and if I don't do it your way, you don't even see my function as
>"seperation".

OK. My net deals strictly with discrete signals or pulses. This analog
business is irrelevant since analog signals are easily converted to
discrete pulses. The brain processes discrete changes, period. All
signals in my system are assumed to be complementary, meaning that, if
a signal stands for the onset of a phenomenon, there must be a signal
that represents the offset of the same phenomenon. IOW, a change in
one direction must be followed by a change in the other direction. The
complementarity principles figures prominently in my scheme. Signals
arriving from sensors are mixed. They must be separated. Signal
separation is an extremely simple process. I explain how I do it on my
site:

http://www.rebelscience.org/AI/perceptual_network.htm

After separation the signals must be fused using coincidence cells.
This significantly shrinks the problem space.

After fusion, signals are sent to the memory layer where they are
sequenced. The memory layer is the heart of the system. This is where
predictions are made. It consists of a huge number of seven-node
sequences. Each node can store an interval or trace. Knowing one
interval is enough to deduce the others. This is a must for pattern
completion.

The nodes send their outputs to the motor output layer via the motor
learning layer where effector connections are made.

The behavior or concept formation layer sits on top of memory and
organizes the sequences into coherent behavior groups. The thinking
mechanism can scan memory (at varying speeds), resolve conflcits and
create new concepts automatically.

The motor output layer is the inverse of the sensor layer. Actions
have an onset and an offset. The motor output layer sends conflict
signals (there are two types of motor conflicts) back to the motor
learning layer where the conflicts are resoloved.

This, in a nutshell, is how my network is organized. I am still
working on the concept formation layer. I try to be as biologically
plausible as I can. As you can see, there is very little ressemblance
between your network and mine. This is why we keep talking past each
other.

>As I said, if I'm not doing it your way, you don't think I'm doing it at
>all. You are making a mistake by not seeing that there are many different
>ways to skin the cat here.

I am not one to discourage anybody from their dream but I always tell
it like I see it. IMO, you are wasting your time, big time. Which is
fine with me. I have wasted many years chasing red herrings. Best of
luck.

Louis Savain

Why Software Is Bad and What We Can Do to Fix It:
http://www.rebelscience.org/Cosas/Reliability.htm
.