Machine learning in trading: theory, models, practice and algo-trading - page 1001
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
RL is Reinforcement Learning?
Game models directly related to the market would be interesting. For example, one could try to simulate the process of hedging of traders' skewed positions by brokers. Maybe there are some stable patterns of price behavior (due to the inevitable time lag between the accumulation of skewness and its hedging). Although, everything must have been calculated a long time ago.
The English article doesn't have Mandelbrot at all for some reason. I can add it there.)
Yes. If you can get this information somewhere why not, or at least the history of the sentiment from the oanda. The combinatorial had it. Why have not entered Benoit I do not know, there is obviously a fractal there and he is the creator :)
Am I correct in assuming that this is Combinator:https://www.mql5.com/ru/users/thexpert?
I suspect that the article was written by a mathematician, and they seem to dislike him a bit)
Am I correct in assuming that this is Combinator:https://www.mql5.com/ru/users/thexpert?
I suspect that the article was written by a mathematician, and they seem to dislike him a bit)
Yes, he did. Maybe he was kind of considered as a dissident.
Thank you.
I don't know exactly. For example, there was some controversy about the priority in the opening of the Mandelbrot set.
The question to the old-timers, has anyone tried to search for levels with mo?
Just like everything else, improvise an algorithm to calculate levels and use it as a target, and you can vary the features to your liking, picking up those that somehow affect the target. In general, of course "levels" will be more complicated than the color of the next bar, that is, simply put, there is nothing, given that the color of the next bar is predicted with an accuracy of 1-2% above random.
The color of the next bar, that is, simply put, nothing at all, given that the color of the next bar is predicted with an accuracy of 1-2% above random.
I nevertheless hope not at all), after a lot of experiments with "MO" I came to a conclusion that the market as "BP" cannot be predicted (I think) because of non-stationarity of "everything" beginning from predictors and finishing with the target system itself, everything is more strict and unambiguous with levels. But how to convert the information correctly I have no idea, of course I can add a few last extrema (a la levels) as predictors but it's too primitive and in fact it will be the same "BP", I would like to think that the "MO" should remember what happened at the current price in the past, and even in a statsionary form
I still hope not), after a huge number of experiments with "MO" I've come to the conclusion that the market in the form of "BP" is impossible to predict (I think) due to non-stationarity of "everything" starting from predictors and finishing with a target, everything is more strict and unambiguous with levels. But how to convert the information correctly I have no idea, of course I can add several last extrema (a la levels) as predictors but it's too primitive and in fact it will be the same "BP", I would like to think that the "MO" should remember what happened at the current price in the past, and even in a statsionary form
Well, the main development is not in the MO, but in the algorithm of "levels". In fact, it should be a sort of a channel, but not just some BB, but taking into account the past extremums and their clustering, if it still remains on the market...
To begin with, it is necessary to verify the very fact that tn. "Levels" exist in the market. That is the fact that past extrema shall somehow influence the future ones, i.e. there should be clustering and if there is clustering, then let's proceed further.
I still hope not), after a huge number of experiments with "MO" I've come to the conclusion that the market in the form of "BP" is impossible to predict (I think) due to non-stationarity of "everything" starting from predictors and finishing with a target, everything is more strict and unambiguous with levels. Of course I can add several last extrema (a la levels) as predictors, but it's too primitive and in fact it will be the same "BP", I wish I had something that the "MO" would remember what had happened at the current price in the past, and even in a statsionary form.
The market remembers everything, only you do not see it yet. This mystery will be revealed by nature's will or persistence. By the way...One does not prevent the other.
The market remembers everything, only you don't see it yet. This mystery will be revealed by nature's will or persistence. By the way...One does not prevent the other.
Thank you for your kind words ...
Driven by the fierce desire to revive this branch, and taking into account that forecasting is possible exclusively and only on stationary VR
(1. Kolmogorov A. N. Interpolation andextrapolationof stationaryrandomsequences
2.Wiener N.Extrapolation, interpolation and smoothing of stationary time series)
Question:
In fact, the value CLOSE[i]-OPEN[i] is nothing but the sum of the increments.
A sequence of such values should, in the limit, tend to a normal distribution.
Well, there is an opinion that the sequence of returnees (CLOSE[i]-OPEN[i])-(CLOSE[i-1]-OPEN[i-1]) is a stationary series.
Has anyone tried such a thing on the NS input and what were the results???
P.S. Max, Doc, Mishanya, Koldun, Alyosha... Who are you throwing this thread at? А?