Machine learning in trading: theory, models, practice and algo-trading - page 1442
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
Mandelbrot has a useful picture about prices, where each of their movements is represented as three (the second is the maximum correction) and a multifractal price structure is based on this. This can be written as a rule of formal grammar: T -> TCT
However, I do not see much sense in this approach. Mainly because of the same non-stationarity. So there is a lot of room for theorists there - stochastic grammars, etc.)
Yes, yes, I've seen these pictures of "oscillators". Essentially an expanded representation of Elliott waves (the book is lying around too). If you take some simple oscillator-pattern as a basis, then you can describe a market fractal, which, as you note, will then die because of non-stationarity.
I don't know about the formal grammar, I can`t do it today, tomorrow maybe.
Here I exported some alphabet or priceaction or inter-bar sequences to csv, 2 files to search for 4 connections and 5
I gave up doing that a long time ago, but I've made sure that it doesn't matter what algorithm is used to find information between bars/fractals etc. - statistically there will always be similar results ;)
in curly brackets this is the alphabet you're looking for.... i.e. i gave up looking for the alphabet a long time ago - it is not in bars!
Time is not the best way to describe CD - if there is no interest of participants in the asset, there will be some fluctuations of CD caused by algorithms of market makers of price formation - i.e. the information will always have distortions
Just to code the sequence - I've looked, there are no repeats, or rather about 50% of OHLC will have repeats of combinations of candlesticks, and the remaining 50% will have an equal number of statically insignificant combinations
I.e. in numbers: here is the history of 132623 bars H1, my script finds 14181 repeats of combinations of 4 bars on the history, here is the statistics of repeats of these patterns? :
and then there will be a gradual decrease in the number of repetitions found by 10-20 pieces, and so on up to 30-40 pieces, then at the end there will be 1-2 pieces - i.e. no alphabet)),
And it doesn't matter how many candlesticks you take for analysis - even 3, even 5, even 10 - you will always get the following statistics - first there will be a lot of detected patterns, and then the number of patterns will evenly decrease, i.e. it will be a statistically inclined bell;)
It is only a matter of principle possibility to encode the price (tick) by finite words from the finite alphabet. This possibility arises because of the discreteness of time and price.
In practice, in this particular language we will have long sequences of moments without change, only occasionally interspersed with ticks. There is no practical sense in it.
There is no practical sense in it.
I've added files for H1, you can generate them for any TF, but found stat. repeats have no information - they don't alternate in a certain sequence - i.e. they don't add words or information, they are just statistically there!
Here I exported some alphabet or priceaction or inter-bar sequences to csv, 2 files to search for 4 connections and 5
I gave up doing that a long time ago, but I've made sure that it doesn't matter what algorithm is used to find information between bars/fractals etc. - statistically there will always be similar results ;)
in curly brackets this is the alphabet you're looking for.... i.e. i gave up looking for the alphabet a long time ago - it is not in bars!
That's what I'm talking about - alphabet, language, grammar, etc. can be entered for SB. But if they do not help to find price differences from the SB, then there is no sense in them).
That's what I'm talking about - alphabet, language, grammar, etc. can be introduced for SB. But if they do not help to find differences between prices and SB, then there is no point in them).
I don't remember all higher mathematics, although I've always had a well-deserved "5" ((((, but I'm saying that no matter how "twist" any TF is, even ticks, even though they've reached alphabets, but statistically there will always be such a picture:
All this forum is full of such graphs and it doesn't matter what and how authors of their methods searched, and any use of statistics for TS will have a gain for the market ... If you try to use any statistics you'll win the market, but if you'll realize that the market is always winning, then you'll probably get the best result (depending on the time of testing, a year, two years, 10 years))) All profitable TS come out with profits only due to successfully chosen money management and certain fortunate coincidences))
I don't remember all higher mathematics, even though I've always had my 5 (((, but I'm saying that no matter how "twist" any TF is, even tics, even alphabets have been reached, but statistically there will always be such a picture:
This is the way this whole forum is filled with charts, and it doesn't matter what the authors of their methods were looking for (((
When I look at the calendar API in mt5 it unfortunately does not work in the Strategy Tester. I wanted to make my charts with %rates and other fundamental info. It should be interesting.
The charts themselves do not reflect the real change in the value of a currency, only in relation to something. For example, to another currency and interest rate / interest rate differential, inflation... whatever.
There are plenty of possibilities for transformation by including substantial information and not only differentiation
Yes, I thought about it, last month I had an interesting job - I developed an online economic calendar, the customer has been using analytics for quite a long time and judging by the payment he did not feel sorry for the money spent
i.e. maybe the pure technical analysis (without news analysis) makes no sense
fundamentals fast
It's interesting to use bonds as a leading indicator
Foundation First
with bonds is also interesting as a leading indicator
what is the GMT of the metaquotes server?
what is the GMT of the metaquot server?
I don't know, you can look through TimeCurrent(), TimeGMT()