Discussion of article "Grokking market "memory" through differentiation and entropy analysis"
I usually read articles diagonally, but I read this one slowly and in its entirety. Very curious system of "digesting" market information demonstrated by the author.
Thank you for an interesting topic and implementation.
At the same interval, the tester showed a different picture. But I would never take the tester's picture as adequate in this case.
It should be well understood that the tester shows complete bullshit at market entries in the mode at opening prices.
Well, how can one treat one's labour in such a mocking way, when at the last stage of application one's head crashes on the rocks like this!
Regarding Article to *** - bar closing prices (no, I'm not advocating ticks).
Liquid substance - because wild loss of information.
It is quite possible that there was a theoretical grail in hand, but such a practical application of the theory can hardly show it even on backtests.
For self-potimisation a fat plus to the Author. The article is cool.
A quick read.
I disagree with some things - because the article does not pay due attention to the methods of determining the sample size to work with.
But, in general, the research direction is promising.
Entropy is the most important parameter for determining market persistence, and in physics it is given special attention. Moreover, we can say that in physics, the entropy (non-entropy) of a process is the only coefficient that carries deep meaning and gives a qualitative understanding of randomness/determinism of events. It is a serious alternative to autocorrelation coefficients, Hirst coefficients, etc. when used as a so-called "indicator of discord".
fxsaber:
As for the Article before the Tester, it is applied to bar closing prices (no, I am not in favour of ticks).
It is quite possible that there was a theoretical grail in hand, but such a practical application of the theory is unlikely to show it even on backtests.
I completely agree. CLOSE/OPEN M1, M5, .... is no help to us. But transition to ticks and their thinning is a delicate thing, you should know how to work with it.
In general, Max has a lot more to do, but as a first swallow, the article is good.
Anyway, Max has a lot more to come, but as a first swallow, the article is a good one.
The article is great in terms of the style of presentation and the material in it. It was very pleasant to feel like a complete zero even after trying to understand the algorithms of the source...
However, there are always connoisseurs of "goodness" who have deep knowledge to evaluate the endeavours of "younger comrades".
It's a pity that paths with such super-specialists cannot but cross. To the non-existent black list, in short.
and there will be interesting results on opening prices.
The paper has two applications.
One is theoretical. There, indeed, anything can be substituted.
The other is practical: Tester. You can't do that. You have simply destroyed expectation by this method.
But it is through the Tester that you analyse prospects. And it shows such a picture that you will throw a lot of good ideas into the basket.
ZЫ It is necessary to have some kind of competent instruction on the use of the Tester.
The article is great in terms of the style of presentation and the material in it. It was very pleasant to feel like a complete zero even after trying to understand the algorithms of the source...
However, there are always connoisseurs of "goodness" who have deep knowledge to evaluate the efforts of "younger comrades".
It's a pity that paths with such super-specialists cannot but cross. To the non-existent black list, in short.
I've got nothing to talk to you about, mate. If you don't know physics, go away, Vasya, and don't cough.
I'm talking about the article - it is the first attempt to study the entropy of the market, and rightly so.
There are two applications in the article.
One is theoretical. There, indeed, anything can be substituted.
The other is practical: Tester. You can't do that. You just destroyed the expectation matrix with this method.
But it is through the Tester that you analyse prospects. And it shows such a picture that you will end up throwing a lot of good ideas into the basket.
I'm talking about the article - it attempts to study market entropy for the first time and rightly so.
The results are on the table!
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
New article Grokking market "memory" through differentiation and entropy analysis has been published:
The scope of use of fractional differentiation is wide enough. For example, a differentiated series is usually input into machine learning algorithms. The problem is that it is necessary to display new data in accordance with the available history, which the machine learning model can recognize. In this article we will consider an original approach to time series differentiation. The article additionally contains an example of a self optimizing trading system based on a received differentiated series.
The Expert Advisor was run with the specified hyperparameters without genetic optimization, i.e. almost at random, on the EURUSD pair with the 15-minute timeframe, at Open prices.
Fig. 5. Tested Expert Advisor settings
Fig. 6. Results of testing with the specified settings
Fig. 7. Virtual tester results in the training sample
In this interval, the implementation showed a stable growth, which means that the approach can be interesting for further analysis.
Author: Maxim Dmitrievsky