Machine learning in trading: theory, models, practice and algo-trading - page 1584
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
Not really. Igor is very correct in what he says. I realized a long time ago that we should separate the trading strategy into parts. To put it simply.
Primary - this is how we define the situation when to enter and when to exit.
Secondary - this is how we enter and exit, ... averaging, partial closing, maintenance in short.
The primary one - a minimum of parameters and maximum simplicity for stability. We use it to put all of our understanding of the market and all types of analysis and experience.
The secondary - the market has nothing to do with it, there is a pure game theory of maximizing one's winnings. The main thing is not to change the parameters of the primary one.
Primary and secondary to adjust and optimize and analyze separately, I hope I explained the idea clearly.
Well, the "tertiary" .... is the main unresolved problem .... determine when the strategy stopped working, and from the word stopped working at all - in this direction I see only a variant of the formation of trading statistics online on the principle of the Strategy Tester report, to make it all work with the first transaction in the test need to take part of the transactions from the tester, then it will be possible with a significant deviation from the tester report to turn off EA.... all in all, i have a lot of work to do (((
well and "tertiary" .... the main unresolved problem .... determine when the strategy stopped working, and from the word stopped working at all - in this direction, I see only a variant of the formation of trade statistics online on the principle of the Strategy Tester report, to make it all work with the first transaction in the test need to take part of the transactions from the tester, then it will be possible at a significant deviation from the tester report to turn off EA.... All in all, there is a lot more work to write and study (((
Exactly, by the way, I'm dividing secondary into 3-7 sub-strategies now. In case many didn't understand, I put the same question in that very topic - how to make sure the system is not working
https://www.mql5.com/ru/forum/329200
Not really. Igor is very correct in what he says. I realized a long time ago that we should separate the trading strategy into parts. To put it simply.
Primary - this is how we define the situation when to enter and when to exit.
Secondary - this is how we enter and exit, ... averaging, partial closing, maintenance in short.
The primary one - a minimum of parameters and maximum simplicity for stability. We use it to put all of our understanding of the market and all types of analysis and experience.
The secondary - the market has nothing to do with it, there is a pure game theory to maximize your winnings. The main thing is not to change the parameters of the primary one.
Primary and secondary to adjust and optimize and analyze separately, I hope I explained the idea clearly.
Or the third way is to guess the extremums and open a position on them, and therefore we should always be in the market. Then we don't need to divide into the primary and secondary parts!
Or the third way, to guess extrema and open a position on them, those always be in the market, then divide them into a primary and a secondary part, you don't need it!
Well, if you can guess, please. The drawdown of each deal should be zero. Good luck! ;) and we're talking about real stuff here.
Well, if you can guess, please. Open on the peak tick on the whole cutlet, the drawdown of each transaction should be zero. Good luck! ;) and here we are talking about real things like.
I will not interfere ...)), ok, speak up, I will not interfere ...
Sorry for the intrusion
Not really. Igor is very correct in what he says. I realized a long time ago that we should separate the trading strategy into parts. To put it simply.
Primary - this is how we define the situation when to enter and when to exit.
Secondary is how we enter and exit, ... averaging, fractional, partial closing, maintenance in short.
The primary one - a minimum of parameters and maximum simplicity for stability. We use it to put all of our understanding of the market and all types of analysis and experience.
The secondary - the market has nothing to do with it, there is a pure game theory to maximize your winnings. The main thing is not to touch the parameters of the primary one.
The primary and secondary ones should be adjusted and optimized and analyzed separately, I hope I explained the idea clearly.
I hope I have explained my thought clearly.
I am writing about search for regularities, he makes some kind of molding in the form of martingale without any explanation.
Let's avoid all this mess, the main thing in predicting BP is patterns (read: repeating atomic events that can be predicted). If you have something to say on the subject, I will listen and even pretend to sympathize.
The conversation started with this, then he turned on the dope for some reason. Some kind of verbal diarrhea.I've been talking to him for a while now.
I write about the search for patterns, he throws some kind of molding in the form of martingale, without explanation
Let's avoid all this mess, the main thing in predicting BP is patterns (read: repeating atomic events that can be predicted). If you have anything on the subject, I'll listen and even pretend to sympathize.
Max, let me ask you a question... How will you solve the problem of the market fractality? We all know that the market seems to have regularities! Today the "head and shoulders (for example)" was forming for 15 min, and yesterday it took us 6 hours.
So how will you teach the MO algorithm using non-repeating data in the future) and how will you trade patterns that will not repeat in the future?)
differentiation is not useful here))
Max, let me ask you a question... How will you solve the problem of fractality of the market? We all know that the market seems to have regularities! Today the "head and shoulders (for example)" was formed for 15 minutes and yesterday it took us 6 hours.
So how will you teach the MO algorithm using non-repeating data in the future) and how will you trade patterns that will not repeat in the future?
differentiation will not help))
Fractal does not mean predictability (see Mandelbrot), literally fractus - lat. fractional, jagged.
There is no martingale there, not at all, there is a simple placing of orders, and averaging and counter orders, there is reopening of orders with movement along the trajectories, if "to tell on the fingers" - this is a search for grids of orders with a variable number of allowed orders, but GA itself moves orders according to a principle known to him, well almost )))) But this is not the main aim of searching for an efficient TS, this is just to find out the possibilities of the GA strategy tester and analyze the options of working with orders
Just write an article and stop worrying about it, I'm not a psychic!
fractal does not mean predictable (see Mandelbrot), literally fractus is Latin for fractional, jagged.
Of course, that's not what I mean....
Roughly speaking, there is a pattern, a graphic one, from the price... It repeats(fractality), but always with different parameters, speaking the language of DSP the pattern always has different amplitude and frequency, so it turns out that this pattern never repeats )))) if we analyze the market with a sliding window and indicators with fixed (non-adaptive parameters)
This is a problem, it should be solved before solving all the following ones