[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Artificial Intelligence, expert and neural



PureBytes Links

Trading Reference Links

John!
The term heteroskedasticity refers to the distribution of the variances
(or standard deviations). Simply put in this case the variances are not
constant as with a uniform normal distribution but vary. In the case of
autoregressive heteroskedasticity the current variance is dependent on
the previous variances either positively or negatively. If the
autocorrelation among the variances is positive then high past variances
increase the likelihood of a currently high variance and vice versa.
If the autocorrelation is negative then high past variances increase
the likelihood of a currently low variance.
Regards,
Yngvi
hardy@xxxxxxxxxxxxx

-----Original Message-----
From: owner-metastock@xxxxxxxxxxxxx
[mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of jhmtn
Sent: Wednesday, February 07, 2001 10:19
To: metastock@xxxxxxxxxxxxx
Subject: Re: Artificial Intelligence, expert and neural


David,

My word!  Its been many (too many) years since I heard the term
'heteroskedasticity' ---- buried back there somewhere in that lesson
I must have slept through in my college statistics class!

It has to do with measuring autocorrelation in a data time series doesn't
it?

Curious? ........ John

----- Original Message -----
From: "David Jennings" <davidjennings@xxxxxxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: Wednesday, February 07, 2001 1:40 AM
Subject: Re: Artificial Intelligence, expert and neural


> (Generalised) Autoregressive Conditional Heterskedasticity - these are
> basically non linear fat tailed stochastic processes with a high peaked
> frequency distribution. Most market have statistically significant ARCH.
>
> GA - genetic algorithm     NN neural network.
>
> DJ
> ----- Original Message -----
> From: "Lionel Issen" <lissen@xxxxxxxxx>
> To: <metastock@xxxxxxxxxxxxx>
> Sent: Wednesday, February 07, 2001 4:01 AM
> Subject: Re: Artificial Intelligence, expert and neural
>
>
> > David:
> > Please clarify the Terminology for me: GARCH, ARCH, GA, NN
> >
> > If the team from Rice U were successful, they should be reticent. Since
> they
> > aren't selling software, books or courses, I suspect that they are
> > successful.
> >
> > The book "The Predictors", was too boastful for my tastes.
> >
> > Thanks for your input.
> > Lionel Issen
> >
> > ----- Original Message -----
> > From: "David Jennings" <davidjennings@xxxxxxxxxxxxx>
> > To: <metastock@xxxxxxxxxxxxx>
> > Sent: Tuesday, February 06, 2001 4:51 PM
> > Subject: Re: Artificial Intelligence, expert and neural
> >
> >
> > > There are quite a few instances of success trading with either GAs or
> NNs.
> > I
> > > must confess, I found The Predictors a bit populist. I'd hoped to get
> some
> > > insighst into model structure. Quite a number of the banks put up big
> > money
> > > for research into fuzzy networks, GARCH and ARCH. $5M p.a. was a
typical
> > > research budget.
> > >
> > > The ultimate Trading Guide is worth a read on how good GAs and NNs
are.
> I
> > > must admit I'm a great fan of GAs for seeking optimal parameters for
an
> > > indicator over a portfolio. Model structure is very important for
> trading
> > a
> > > market - whatever chaos theory tells us, and thus one can end up with
an
> > > overly complex model which misses the point.
> > >
> > > My experience of NNs is that one has to be quite specific about one's
> > target
> > > e.g. predicting turning points etc.and then use them in combination.
> I'm
> > > convinced that a guy with sound experience in say Fib could do as good
a
> > job
> > > as any NN. however, Lionel makes a strong point though, the beauty of
AI
> > in
> > > general is that it can play multiple markets in multiple time frames
> > > simultaneously - and hence generate super normal gains or losses.
There
> > was
> > > a group of guys from Rice university (whose names I've forgotten) who
> > > launched Frontier specialising in pattern recognition. They built
> trading
> > > systems and played the markets. The very absence of research papers
from
> > > them suggests that they were successful!
> > >
> > > DJ
> > > ----- Original Message -----
> > > From: "Lionel Issen" <lissen@xxxxxxxxx>
> > > To: <metastock@xxxxxxxxxxxxx>
> > > Sent: Tuesday, February 06, 2001 8:03 PM
> > > Subject: Re: Artificial Intelligence, expert and neural
> > >
> > >
> > > > Jeff:
> > > > A few years ago I read that the Bankers Trust of  NY was spending a
> mint
> > > on
> > > > such things like neural nets, artificial intelligence and chaos
> systems.
> > > > Last year the Deusche Bank bought them out and like many foreign
> > companies
> > > > they are very secretive, and wont give an outsider the time of day.
So
> > > there
> > > > is no way of finding out how successful these methods are.While
> current
> > > > daily prices are available on the net,  charts are not.
> > > >
> > > > There was a recent book "The Predictors" about a group in the
> southwest
> > > (New
> > > > Mexico?) who were/are successful at this trading in esoteric
> > instruments.
> > > My
> > > > feeling is that if they were very successful, they'd make someone
like
> > > > Buffett look like a penny ante player.
> > > > Lionel Issen
> > > > lissen@xxxxxxxxxxxxxx
> > > > ----- Original Message -----
> > > > From: "Jeff Haferman" <haferman@xxxxxxxxxxxxxxxxxxxxxxx>
> > > > To: <metastock@xxxxxxxxxxxxx>
> > > > Sent: Tuesday, February 06, 2001 11:51 AM
> > > > Subject: Re: Artificial Intelligence, expert and neural
> > > >
> > > >
> > > > >
> > > > >
> http://www.emsl.pnl.gov:2080/proj/neuron/neural/systems/shareware.html
> > > > >
> > > > > has quite a selection.   Neuroforecaster/GA  is geared
specifically
> > > > > toward financial applications, and you can download a limited
> > > > > version.  The shareware neural net package SNNS probably has the
> > > > > best reputation.
> > > > >
> > > > > But you're going to spend a lot of time training and tuning any
> > > > > expert system or neural net.  The results I've seen indicate that
> > > > > neural nets are great at learning about past market activity,
> > > > > but I haven't seen any convincing results that show that
> > > > > they do better than random at predicting future activity.
> > > > >
> > > > > Jos Carlos Duarte Areia wrote:
> > > > > >I'm  looking for this subject, anyone knows the best sites and
> > > softwares
> > > > (I
> > > > > >found any at TASC 2000 page 19, bonus issue)to look for ?
> > > > > >Thanks JCarlos
> > > > >
> > >
> >_________________________________________________________________________
> > > > > >Get Your Private, Free E-mail from MSN Hotmail at
> > > http://www.hotmail.com.
> > > > > >
> > > >
> > > >
> > >
> >
> >
>