NATURAL INTELLIGENCE as the basis of a trading system
The short answer is Yes. For the long answer, see Schwager's book The New Market Wizards. Conversations with the best traders in America. Part three. The silence of the "Turtles".
Rosh:
The short answer is Yes. For the long answer, see Schwager's book The New Market Wizards. Conversations with the best traders in America. Part three. Silence of the "Turtles".
The book, as I understand it, is not free.The short answer is Yes. For the long answer, see Schwager's book The New Market Wizards. Conversations with the best traders in America. Part three. Silence of the "Turtles".
Please outline at least the general points of this chapter.
Implex:
Please outline at least the general points of this chapter.
What a request...
Rosh:
The short answer is Yes. For the long answer, see Schwager's book New Market Wizards. Conversations with America's best traders. Part three. The Silence of the Turtles.
The book, as I understand it, is not free.The short answer is Yes. For the long answer, see Schwager's book New Market Wizards. Conversations with America's best traders. Part three. The Silence of the Turtles.
Please outline at least the general points of this chapter.
Unfortunately (or vice versa :), neural networks have not found a worthy application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is still useful to read books :)
So let's wait for five-level developments :)
But it is still useful to read books :)
rebus:
Unfortunately (or vice versa :), neural networks have not found a worthy application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is still useful to read books :)
The question was Unfortunately (or vice versa :), neural networks have not found a worthy application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is still useful to read books :)
Is it possible to train the brain (I say exaggeratedly) on historical data for predictive purposes? That is, learning what many programmers and mathematicians try to teach the ANN. Learning by the same methods (but on a different plane).
Rosh:
Possibilities of a brain are used only on unity percents, besides there are such areas as a subconscious and a superconscious, and they are almost not used, at least consciously by the majority of people. rebus:
Unfortunately (or vice versa :), neural networks have not found a decent application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is still useful to read books :)
The question was Unfortunately (or vice versa :), neural networks have not found a decent application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is still useful to read books :)
Is it possible to train the brain (I say exaggeratedly) on historical data for predictive purposes? That is, learning what many programmers and mathematicians try to teach the ANN. Training with the same methods (but on a different plane).
This question does not have a correct aim, one can teach brain to do many things, including forecasting at Forex market, its abilities are huge, but do not try to do it with methods that are used to train neural networks, nothing good will come out.
rebus:
Unfortunately (or vice versa :), neural networks have not found a worthy application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is useful to read books :)
There was a mathematician Turing.Unfortunately (or vice versa :), neural networks have not found a worthy application for a very simple reason. The fact is that recent research has shown that the human brain neuron works according to a five-level logic. Earlier it was believed (and it formed a basis for development of NS) that logic is two-level. But only spinal cord works with two-level logic. As a result, all hitherto designed computers will forever remain at the level of spinal cord intelligence. The same applies to two-level NS.
So let's wait for five-level developments :)
But it is useful to read books :)
He had invented such a machine - Turing machine.
It is a simple device with a memory ribbon and primitive commands (it cannot get any simpler).
And he proved Turing's theorem: "An algorithm of any complexity can be implemented on the Turing machine".
I mean the phrase - "As a result, all hitherto designed computers will forever remain at the level of spinal-brain intelligence. The same applies to two-level NS." - contradicts Turing's theorem, and is therefore incorrect :))
In order to train a "brain", you have to have one.
Rebus- what is meant by "two-level logic"? That the neuron either works or it doesn't?
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Undoubtedly, this is one of the most progressive and promising directions in science (despite numerous drops of interest in it due to subjective and objective reasons).
However, to date, neural networks have not achieved acceptable results in many tasks. Most likely this is caused by a lack of machines with sufficient performance and specific abilities, but this is a huge separate topic, which cannot and should not be discussed in this forum.
Based on the above, I would like to pay special attention to natural neural networks (brains).
Many people trying to create an EA or other prediction apparatus based on artificial neural networks explain their choice (ANN) by the insufficient ability of the human brain to store large amounts of data for subsequent analysis. However they most likely do not take into account such fact, as performance of that machine (most likely computer), on which they implement this machine. Failure to take into account this critical factor reduces all algorithmic possibilities to nothing. Since in the brain computing takes place in parallel, while creation of a machine with parallel computing (although it is possible nowadays) requires a lot of different expenses, which most likely are beyond a reach of most "ordinary" people, implementation of a neural network of a sufficient power on a computer with sequential calculations is impossible from the practical point of view (theoretically such a network can certainly be implemented, however time spent on training it on "productive" computers may go far beyond the limit).
Since this forum, according to its name, discusses trading strategies and their testing, the subject I suggest will be interesting not only for me (IMHO).
My question is formulated as follows:
Is it possible to train the brain (I say exaggeratedly) on historical data for predictive purposes? That is, training what many programmers and mathematicians try to teach the ANN. Training with the same methods (but on a different plane).
Thank you.