Kohonen and Patterns - page 4

 
ANG3110 писал (а) >>

Hi! You'd better tell me a little bit about how you're doing?

I'm fine. Working...

 
TheXpert писал (а) >>

You have not fully understood me. However, I didn't make my point fully, so I'm correcting my mistake.

Suppose there is a network. There is a TS which trades according to the network prediction. The network predicts daily rates for 5 days ahead.

So. In order to increase the effectiveness of TS we may introduce estimation of forecasts not on history, but in real trading. On this assessment can be tied MM, etc.

It is naturally possible, but long. The tester gives the result very close to reality. And you can make an estimate quickly and immediately. I take the view that all battles are won before they even begin. This also applies to trading, as long as you don't fall into the illusion of over-optimisation.

If you mean the current control of coincidence and divergence, then it is of course very important for Expert Advisors, because I have not got to it yet, as the simple program routine takes much time. While making network, while processing data, while displaying on chart and so on and so forth...

 
ANG3110 писал (а) >>

It is naturally possible, but it takes a long time. The tester gives a result very close to reality. And you can make an assessment quickly and immediately. I take the view that all battles are won before they start. This also applies to trading, as long as you don't fall into the illusion of over-optimisation.

If you mean the current control of coincidence and divergence, then it is of course very important for Expert Advisors, because I have not got to it yet, as the simple program routine takes much time. While making network, while processing data, while displaying on chart and so on and so forth...

Neither.

 
TheXpert писал (а) >>

So. To increase the effectiveness of the TS, you can introduce forecast evaluation, not on the history, but in real trading. This estimation can be used as a basis for MM, etc.

What do you mean by prediction estimation in real trading? It's not quite clear. If you are really interested in my answer, then state your opinion in a bit more detail.

 
ANG3110 писал (а) >>

What do you mean by the estimate of the pronose in a real trade? That's not really clear. If you are really interested in me answering your opinion, then state it in a little more detail.

That if you are using exactly what I think you are using for forecasting, then there is the option of a fairly accurate estimate of the network forecast at once when forecasting.

 
I used a General Regression Network (GRNN) to build the above pictures. This is a modification of the probabilistic PNN designed for approximations and predictions. What's a variant of a fairly accurate estimate of the network prediction right off the bat in prediction? Describe something, you're already on page 2 hinting at something, but haven't written a single word on the merits. I've worked with almost every type of network in the world and I don't think I'll have any trouble understanding what you're talking about.
 
ANG3110 >> :

A 3 month forward forecast for the pound on the Daily, for the duration of the Championship.

-

Look, the pound is dancing exactly as predicted so far!

 
muallch писал(а) >>

Look, the pound is dancing exactly as predicted!

>> so far, so good.

 

That's the way to make it clearer!

 
ANG3110 if it's no secret how did you filter the data or what?