Using Neural Networks in Trading. - page 18
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
cool... let's not go into how you get a trend curve from regression lines... I can roughly imagine that (although frankly I have my doubts about the adequacy of such method, but okay)...
but even the graph shows that the variance of the random term (the noise component) is quite large in some areas... Anyway, I can hardly believe that it's even bigger on smaller timeframes... although maybe it's just some kind of technology feature...
Anyway, I can't imagine that an ordinary linear regression would have a bigger dispersion on small timeframes than on larger ones...
Was this about variance? I meant price prediction. It's easier to predict the daily or weekly price than the minute or hourly, in my opinion
Was I talking about variance? I meant price prediction. It's easier to predict a daily or weekly price than a minute or hourly price, in my opinion
no, well i mean you are predicting - and there is some prediction error - the inconsistency of the regression... you can see from the graph that it's quite large in some areas... Well, in any case, such a configuration on a smaller timeframe should have a smaller error value... although, i don't know, you know better... maybe I just don't understand the subtleties of technology...
No, I mean you predict - and there is some kind of prediction error - the regression inconsistency... you can see from the chart that in some sections it is quite large... Well, in any case, such a configuration on a smaller timeframe should have a smaller error value... although, i don't know, you know better... maybe I just don't understand the subtleties of technology...
You don't look at the curve, it serves to identify a possible trend or cloud the mind =) just from my experience, the higher the tf the more stable the TS
You don't look at the curve, it serves to identify a possible trend or cloud the mind =) just from my experience, the higher the tf the more stable the TS
If this curve is only to fog the mind, then of course there is no need to talk about variance... :)
Neural Networks on the forum, is there a specialist in Probabilistic Networks (RBF,PNN) ? There are a couple of questions:
1. I once read somewhere that VNS can be replaced by MLP and the system will work the same only that the operating time and learning time will be swapped. Is it really so?
( Once I myself started with learning on MLP and even wrote my GA for advanced training in real time, but I was not satisfied with input/output search TIME, VNS memorizes new data on every bar)
2. Who ever used multiple nets for prediction and what was the result?
(I've added pnn to indicator may use at least 10 nets at once, I wish the computer gets too hot and has to wait so I have no results)
3. VNS + GA in my future plans, who thinks is worth trying to select inputs using GA, who tried in a mega-wizard program, or better still think with your head?
(I am a programmer and do not understand the philosophy of the market (yet), all I do by gut feeling which takes a lot of time, the smart inputs/outputs therefore I have not found yet).
So has anyone figured out this H-volatility thing? h ttps://www.mql5.com/ru/forum/115616/page14#154564
...go through, go through, don't crowd...
So, neural networks are a powerful thing, but not applicable to forex.
I have already written on this forum, but I will say it again, maybe I will save someone a few billion nerve cells and a lot of personal time.
So, neural networks are a powerful thing, but they are not applicable to forex.
The disadvantages of neural networks can be fixed if you know how:
See Why is the neural network retrained?
And if one doesn't know how, i.e. hands grow out of ass or brains are not enough for elementary education, then everything seems inapplicable, by the principle of retarded generalizing demagogy: if it didn't work for me, it means it is inapplicable for others as well.