Machine learning in trading: theory, models, practice and algo-trading - page 2321
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
No, not that at all...
The market is not stationary, algorithms are not trained on it, they die immediately at birth, what they learned in the past will never be repeated in the future...
What if we try to make it stationary.
1) choose "k" major harmonics on the fly and take them as the market model
2) but these harmonics will also "float" over time in frequency, phase, and amplitude
3) we have to find out how to tune them permanently, so each harmonic always has the same frequency, amplitude and phase
If we obtain it, we will obtain the "market model" made of the sum of sinusoids, which is convenient for studying, and the patterns are always repeating because the harmonics are always in the same diapasons
Try it on an artificial series to begin with. I don't think it is realistic. Even if you think up such a transformation, it will most likely lag behind
These convolutions? This is the basis of filtering.
If you stop cycling through the signals and start MO, you can discover a lot.
But I don't want to fight with the burnt-out DSPs.Try it on an artificial row first. I don't think this is realistic. Even if you think up such a transformation, it will most likely lag behind.
The idea is that the market can be described by two sinusoids + the trend is linear or also a sinusoid.
The sinusoids must have the right frequencies relative to each other to get the classic Eliotian model of 5-3-5... and so on in a circle.
Then all TA figures show themselves, head shoulders, double top etc... And there is no magic here
video - shifting phases in harmonicshttps://radikal.ru/video/oVpWCd1Q1pA
If we find a way to shift prices to the same form (with clear parameters), success is practically guaranteed.
Here is even cooler, already with a trend, clearly visible 3-5 structure...
The idea is that the market can be described by two sine waves + the trend is linear or also a sine wave.
The sinusoids must have the right frequencies relative to each other in order to get the classical Eliot model of 5-3-5... and so on in a circle.
Then all TA figures show themselves, head shoulders, double top etc... And there is no magic here
video - shifting phases in harmonicshttps://radikal.ru/video/oVpWCd1Q1pA
If we find a way to shift prices to the same form (with clear parameters), success is practically guaranteed.
Here is even cooler, already with a trend, clearly visible 3-5 structure...
This is from the area of perpetual motion. The structure changes almost every day.
UPD Try the other side, assume that there are several sinusoids that change their period in some range. There is even a ready-made solution of the Hilbert-Huang Transform. From there it will be easier to solve your problem. But I tell you right away, there will be problems on the right edge with any approach.Prices have their own natural "modes" into which they should be decomposed. What is meant is what Mandelbrot depicted when he talked about their fractality.
Fractality of patterns and similarity are different things. Catching the correlation in similarity at different scales is a lot of work) But there is something in it.
Catching the correlation in the similarity at different scales is a lot of work.
For this purpose the notion of multifractal was invented, for which multifractal spectra are considered. In our case it is more correct to speak about stochastic fractality. It turns out something like this.
For this purpose the notion of multifractal is invented, for which multifractal spectra are considered. In our case, moreover, it is more correct to speak about stochastic fractality. It turns out something like this.
Good article. But for the small length of the series. For me, the fractality of series, is the sameness of some rules, or behavior at different scales. And multifractal spectrum... I do not understand how it can be applied.
Good article. But for the small length of the series. For me, the fractality of series, is the sameness of some rules, or behavior at different scales. And multifractal spectrum... I don't understand how it can be applied.
In my opinion, the article is not very good for our purposes, I just chose it as an illustration of an approach that combines multifractality and stochasticity.
Roughly speaking, multifractal = consisting of many fractals and the spectrum is the dimensions of those basis fractals. But we can play around with the notion of "spectrum" and come up with something suitable for us - for example, a function showing the degree of difference from SB at different scales.
This is from the field of perpetual motion engines. The structure changes almost every day for wave makers.
To hell with wave makers, I just showed that two or three sinusoids are enough to describe the whole "mysterious/mysterious" market with all its pattrons, figures and other crap.
And if it's enough, you can create a market model from them and throw out everything else, because it's garbage...
In scientific circles, in order to predict a complex system, a model of the system is created. The model is a simplified representation of the process, but it retains the necessary properties...
A model is studied rather than a real process and it is a model that is predicted rather than a real process, and what we do???? so the result is the same...
We optimize for noise, nothing more, and with zero understanding of the forecasting object...
That's why we should create a market model first, simple and clear, sine wave is a great tool.
assume that there are several sinusoids that change their period in some range.
What's the point? From one chaos you create another chaos, you have to get away from that on the contrary.
But I tell you right away, on the right edge there will be problems with any approach.
So that's the result of your approach.