Predicting the future with Fourier transforms - page 39
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
I doubt that very much.
Just because price is continuous does not mean that it has continuous information about its behaviour in the future, at each new bar.
I agree with you again ))).
But usually, it is a matter of there being some kind of continuous signal in the quotidian, which can show the direction of the quotidian in the future on each new bar. In principle, the application of this Fourier implies this. But can there be a continuous signal in the kotir that shows the direction of kotir movement in the future, on each new bar?
Theoretically, nothing prevents the presence of this signal, or at least nothing prevents us from thinking - that there is this signal. The indicator line is continuous, so there is always some indication. And practically - find some cases in which it is possible to make the actual prediction. I consider the attempts to make the forecast at every bar to be a stupid exercise. It is necessary to determine the points in which the forecast should be made and the method of forecasting.
Not to argue, but to define the concepts. In particular signal.
Okay, let's say the price is continuous (well, don't take weekends/holidays, gaps etc.). But the signal is information. Yes, the price itself has a continuous information about what it is now. But the price change informs where it will probably "go". And is it correct to call something that has only probability as a signal?
1. Strongly doubtful.
2. just because price is continuous does not mean that it has continuous information about its behaviour in the future, at each new bar.
1. this is personal preference :-)
2. Of course, there is no unambiguity. But it's a good premise.
Look at the behaviour of, at least, MA lines with a long period. Do they often abruptly break and go somewhere else? I have never seen anything like that. That is the continuity.
Not to argue, but to define the notions. In particular signal.
Ok, let's say the price is continuous (well, don't take weekends/holidays, gaps, etc.). But the signal is information. Yes, the price itself has a continuous information about what it is now. But the price change will inform where it will probably "go". Is it correct to call something that has only probability as a signal?
Even any conditionally heavy standard MACD line can be predicted with extreme accuracy. Not to mention the prediction of sinusoidal-like lines.
I agree with you.
MA smoothes, not predicts. In fact, so does Fourier. So, smoothing is the prediction?
Smoothing and forecasting are different uses.
To make a prediction, you must first make an analysis. To do this, you need a selective filter.
Then synthesis (prediction) is done. The forecast is extrapolation of those lines after filtering.
It's necessary to determine the points at which to make a prediction and the method of prediction.
We still have to take a "window" of price change to determine the best forecast points, don't we?
As for methods - again, they will all be based on history, on price behaviour. Surely it's statistics, which means we come back to probability again.
It reminds me of quantum physics - where a particle (read signal) is described by a so-called wave function. Its values are not expressed by ordinary numbers, but by complex numbers. For a moving particle (wave), the function is defined at every point in space and varies in time. But the "particle" ("signal") is not at any particular point. It seems to be smeared in space and is present to some extent everywhere at once, somewhere "concentrating" and somewhere "vanishing".
In the case of synthesis, the further we forecast, the lower the probability of prediction.
Even any conditionally heavy line of a standard MACD can be predicted with the highest accuracy. Not to mention the prediction of sinusoidal-like lines.
Quite possible. But the slow/heavy line is easy to predict imho due to its inertia. As for the non-stationary price...
In general, it would be good to see some charts with examples...
In general, it would be good to see some graphs with examples...
The only thing you don't need to do is repeat the same pictures here as well. Type zhunko in the yandex and you'll find plenty of these similar pictures.
I can explain in private if you need to. They'll get the wrong idea here again + risk getting banned again.
I agree with you )))) But we are talking about some kind of signal being extracted from the quote. This is where the questions arise - what is this signal? what does it mean? what is the signal and all that remains after this signal is highlighted? what is this signal?
I, for example, have not been able to isolate anything other than probabilities. But probability, if it differs from 0 and 1, is no longer a guarantee. I.e. let's say a probability of 0.6 for a take means that in 60% the take may work and in 40% the elk.
How do you distinguish between signal and noise?