"New Neural" is an Open Source neural network engine project for the MetaTrader 5 platform. - page 17

New comment
 
papaklass:
Get what the majority needs.

No, we won't get anything. As practice shows, polling by an incompetent majority is ineffective at all.

We need competent consultants, preferably non-programmers. And probably really need to not get a one-sided purely programmer's point of view.

 
TheXpert:

No, we won't get anything. As practice shows, polling the incompetent majority in general is ineffective.

We need competent consultants, preferably non-programmers. And probably really need to not get a one-sided purely programmer's point of view.

Then let everyone remember who he considers a "competent user of NS not a programmer". And send out separate invitations.

I personally remember only Leov. Maybe someone else knows someone on this forum who can be trusted to rack the brains of programmers?

 
It's probably different. Rather, it is the case when you first have to make a product and offer it. And users, all users, and teapots and coffee pots and golden samovars, after getting acquainted with the product, will discover its necessity and usefulness. Up to 5% of traders are likely to use custom-made нS in C++. The rest have an idea of the new product on the level of Wikipedia. Roughly speaking, there is no demand. There is no demand for THIS in the air. THIS must be made, shown and offered. Everything is imho of course, but one thing is certain, if the ns builder is done well, it will be in demand.
 
Renat 2011.10.22 00:17 2011.10.22 00:17:33
One should plan with this possibility in mind.
 
Mischek:

Probably different. It's more a case of making the product and offering it first.

Oh I do not know ... There are several ways to offer the prospects of the NS

-- show the advantage over standard TA tools. In other words, to make a robust profitable Expert Advisor. Will anyone do it? I don't.

-- To simplify the use of the NS to the level of, say, indices. I'm afraid it's unrealistic too.

-- do everything but pick up inputs and outputs and just show the new way of analysis.

-- implement interaction at the neural packet level. That is essentially remove programming, replacing it with metaprogramming.

And users, all users, and teapots and coffee pots and gold samovars, after getting acquainted with the product, will find it necessary and useful.

The product for this must be the level of neuro-packages like NSDT, no way. You may try to go another way - use a fresh product and compare it with FANN bridge for terminal. At least now :) At least there is an echo grid :)

You could try something autoprogrammable, but that's too much, you just can't handle it.

There is no hovering demand for THIS. This must be done, shown and offered. This should be liked.

See point one.

All imho of course, but one thing is certain, if the constructor is made qualitatively, it will be in demand.

If only...

You'd better tell me how you see it from your own vantage point.

There are at least 4-5 different opinions, and they are all different.

Urain:

We should plan with this possibility in mind.

So far, no way. When it is, then we'll see.
 

Yes, the library should be universal, like a Lego constructor - you can assemble anything you want.

Yes, the library should be easy to use with a template Expert Advisor. Easy to use, so much so that it can be used by a non-programmer.

Yes, the library should have a universal interface both on the input and output, so you can connect everything to it, from indicator values to...

In general, it should be a tool like a Swiss knife, but not a magic wand (however much we would like, but it's not possible).

That's why I'm against this question - "In other words, it's impossible to develop a robust profitable Expert Advisor" - it is utopia. Utopia in the sense that not all people on earth can become rich because of the law of conservation of energy.

 
joo:

So I'm against this formulation of the question - "In other words, make a robust profitable EA" - this is utopia.

Utopia is not utopia, but unrealistic, yes. That leaves three points to choose from or a mix :)
 
TheXpert:

You'd better tell me how you see it from your bell tower.

There are at least 4-5 different opinions, all of them multifaceted.



I'm just at the level of wikipedia.

I need the ability to graphically construct ns from something ready-made. I need something that will prepare formalized signals for ns from any indicators.

There are at least 4 or 5 different opinions, and moreover there are different opinions in general.


Not at least, but everything. And if you don't agree on how to start, everything will die out.

Start with something)) With the simplest one. With the simplest stupid ns.

 
Mischek:

Start with something.) The easiest one. With the simplest stupid ns.

No problem. As soon as MK will provide the repository. But very slowly, at least for now :) . I have a more ambitious project:)
 
TheXpert:
I have a more ambitious project :)

Ouch! That's not what I know about?

Are you intriguing, bragging, inviting ? // underline it.

1...101112131415161718192021222324...100
New comment