Now let's be clear, do patterns work? Let's discuss) - page 9
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
Thank you, Yura, for the permission! I'll think about it. In case it doesn't match... or worse!
Of course it cannot be determined, it can only be predicted. One approach is to train the neural network directly to predict the magnitude of future movement before the zigzag turns. You can read about this approach here: http: //forex-pamm.com/lit/for_for_gen.pdf
Thanks, Alexander, for your reply and interesting link. I will look it through.
It turns out that the problem in this formulation is not formalized? Only NS is able to approach in some sense the optimal slicing on the right edge of time series...
It is also possible to make simultaneously N forecasts for 1, 2, ... N bars ahead and generate a zigzag by them, which is much more labor-intensive. From the viewpoint of trading system profitability, only the first Zig is important for a trader, and he should not care about price changes inside the Zig, moreover, about future slice. Therefore, the forecast of intermediate prices is a waste of time.
NS is as one option, other forecasting methods can be used as well. It's just that NS is closer to me, that's why I gave the example of NS.
This is all obvious and understandable.
The problem, of course, is the exponential divergence of forecast accuracy with increasing horizon (N). And given the low forecast accuracy for market-type BPs, it does not make sense to forecast more than one step ahead.
This is all obvious and understandable.
The problem, of course, is the exponential divergence of forecast accuracy with increasing horizon (N). And given the low forecast accuracy for market-type BPs, it doesn't make sense to forecast more than one step ahead.
Have you considered the Lapunov index? Very good stuff. Try it if you haven't already. Get your forecast horizon.
Judging from the branch an admirable result some patterns work.
And you don't judge results by a branch, you judge them by equity. The threads are more populated by those who are not trading, but flubbing.
From the point of view of a profitable trading system, only the first Zig is important to the trader, and price fluctuations within the Zig, and especially future slicing, should not be of concern to him.
The first Zig is only a potential take level. And for a full-fledged TS it is also important to determine the levels of potential losses.
Judging from the thread an admirable result some patterns work.
yes
e.g.
161 fibo level
--
we put a grid on a worked out zigzag and catch the 161 level
is reached statistically very often
yes
e.g.
161 fibo level
--
we put a grid on a worked out zigzag and catch the 161 level
is reached statistically very often
Yes, it is known to be worked out, only after the next extremum.