Machine learning in trading: theory, models, practice and algo-trading - page 2696
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
So it's done for R - somehow through a dll, though I didn't go into details.
I understood that the speech is about the native.
And so yes, you can knock anywhere via dll.
An analogue of python api was made for R, something there didn't work out to upload it to a local market like pypi, I don't know about the rest. Code in interpreted languages runs slowly, probably there is no great sense, i.e. you won't get fast testing :)
Meaning is a relative concept) In my opinion, it's better when something is available and you can choose whether to use it or not) Fast testing won't work, but at least some kind of testing to evaluate the meaning of further fiddling.
Still, I would like to understand the reason why there is no mt-R analogue for python. I am talking about the possibility of launching an interpreter from an mql5 program with the ability to send commands to it and exchange data in both directions. This is convenient, for example, for quick testing of a trained model without distilling it into mql5-code, and in general it is quite a flexible tool. And it seems to be exactly what a fan of "chattering and chattering" wants.
.
If you need it very much you can do it yourself or order it in freelance. Just make a copy of functionality (you can also make a copy of function names, for portability from 1 to another engine) for Python.
Please explain how to work with the loss function
How to use the target function for minimisation?
And second question.
In Matlab the function fminsearch() uses Nelder-Mead algorithm.
This algorithm is not present in ENUM_LOSS_FUNCTION.
Can we count on adding this algorithm?
In my opinion, it is better when something is available and you can choose whether to use it or not) Fast testing will not work, but at least some kind of testing to assess the meaning of further fiddling.
Maybe it's not about security, but about the complicated Py api for arrays.
And they just didn't bother with it.
There's a really painful numpy api there.
It's not boring, it's just unusual. When you get used to it, it becomes a buzzkill. I would vote for such an api of arrays in C-like languages like MQL.
It's not painful, it's just unusual. When you get used to it, it becomes a thrill. I would vote for such an api of arrays in C-like languages like MQL.
No, don't have it in C-like languages ))
Transmission in bytes, there is nothing better.
The rest is all wrappers, with the imagination of the wrapper developer.
Looking for companions on an interesting and exciting journey down unknown paths in search of mysterious predictors/features/signs.
I have a map by which to find them, need hands for setting traps and thinking about the habits and halo of these amazing phenomena.
The way is not easy, long, but fascinating, and I'm sure we won't be left without trophies!
If in doubt - ask questions!
)
Please explain how to work with the loss function
How to use the target function for minimisation?
And the second question.
In Matlab the function fminsearch() uses Nelder-Mead algorithm.
This algorithm is not present in ENUM_LOSS_FUNCTION.
Can we count on adding this algorithm?
The task of training a neural network is to find an algorithm that minimises the error on the training sample, for this purpose a loss function is used. TheLoss method is used to calculate the deviation, which allows you to specify 1 of the 14 types ofENUM_LOSS_FUNCTION enumeration.
The obtained deviation values are then used to refine the parameters of the neural network, this is done using theDerivative method, which calculates the values of the derivative of the activation function and writes it into the passed vector/matrix.
Due to the ability to extend the enumerations, we can add new algorithms as needed.