Machine learning in trading: theory, models, practice and algo-trading - page 2702

 
Aleksey Nikolayev #:

He says that linear regression is better) And it is much easier to use it in MT than all the fancy stuff).

It's interesting to listen to others, but you have to trust yourself
 
Aleksey Nikolayev #:

Well, if it helps you to make a profit at the same time, why not? Or is your option "I'd rather have a cow die than my neighbour have two"?

Try to use the time spent on trolling to write at least one indicator.

 
Maxim Kuznetsov #:

you got banned from both Google and Yandex at the same time ???

https://rpy2.github.io/

It's more likely that you got banned by the sense analyser. It's about linking MT with python, when from mql program you can run it, send commands and exchange data both ways.

 
Evgeny Dyuka #:

Try to use the time spent on trolling to write at least one indicator.

Indicators are not necessary.

 
Aleksey Nikolayev #:

Indicators are not needed.

)))) I was also puzzled at first, why visualisation of averages)))))

 
Evgeny Dyuka #:

Why waste time and effort on the development of MQL5. This is a private commercial company created to make profit.

According to my modest experience of communication here. It is much easier to find partners for solving tasks by setting specific tasks and giving some perspective on why their solution is needed or a plan of tasks. Joint gathering of participants for much better tasks is difficult, and here individual races, it is like a team of runners to gather and say let's train together, exchange training plans)))).

 
mytarmailS #:
Looking for those who will go with me, but where to go, what to do I'm not ready to reveal....
Alexei, you have a contradiction on contradiction...

You confuse "what to do" with "how to do".

If a person writes that he has not understood anything, then it is necessary to make additional efforts to simplify explanations, and why - you will not do anything together with me anyway, and discuss constructively the same. You already have your own algorithm, which is the best and correct for you.

Well, ask a clarifying question on the merits - I will try to answer.

 
Aleksey Vyazmikin #:

You're confusing "what to do" with "how to do".

If a person writes that he has not understood anything, then it is necessary to make additional efforts to simplify explanations, and why - you will not do anything with me anyway, and you will not discuss constructively either. You already have your own algorithm, which is the best and correct for you.

Well, ask a clarifying question on the merits - I'll try to answer.

It's pointless, all this talk....
As an option you can create a dataset and compete in modelling, it would be interesting, maybe I would finally deploy my model
 
mytarmailS #:
It's pointless, all this talk....
Alternatively you could create a dataset and compete in modelling, that would be interesting, maybe I'd finally deploy a model of mine

There you go, that's what I said, it's pathetic to understand another person.

To compete - yes you can. There are quotes from MQ for MT5, you just need to decide on the signal for decision making - the signal on which the information for model calculation will be transmitted. I have not learnt anything on Forex - predictors all under Si were made - it will be funny to look at their universality.

The winner will be determined in half a year based on trading results.

 
Aleksey Vyazmikin #:

There you go, that's what I said - to understand the other is a pathetic endeavour.

It is possible to compete. There are quotes from MQ for MT5, you just need to decide on the signal for making a decision - the signal on which the information will be transmitted to calculate the model. I didn't learn anything on Forex - predictors were all made for Si - it will be funny to look at their universality.

The winner will be determined in half a year based on trading results.

Isn't it possible to test it on the test? Do you have to trade for half a year?
Reason: