Machine learning in trading: theory, models, practice and algo-trading - page 1931

 
fxsaber:

A question from a beginner.


There are three variables A, B, C. Some kind of condition is made up of them by hand. For example.


I want to reproduce this condition automatically. I don't need to find it, because I already know it. But I need to have for example a dozen of some weight coefficients, a certain combination of which is able to hit this condition with high probability, when I set A, B, C there (polynomial or HC - I don't know, because it's zero) and get the original condition.


I am interested in what kind and how many input weight coefficients the desired function has, so that such original conditions can be reproduced through weights?

you can calculate the result of the condition for ranges A,B,C
and see
red is true
blue is false


 
Vladimir Suslov:

you can calculate the result of the condition for ranges A,B,C
and see
red - true
blue - false


The problem from fxsaber is not quite clear. He proposed three variables in arbitrary condition and apparently asked if this condition can be generated automatically if we consider values of the variables as weights within some system which has a regularity of changes in its parameters.

Alternatively, to synthesize the condition automatically we need:

1. Separate the system as a complex of parameters from the environment of objects.

2. to analyze interdependences of changes inside this complex for a long time, focusing attention at various events of the environment.

3. In the process of observation to fix conditions - i.e. to build relations of values of parameters and events of the environment.

4. To classify obtained conditions.

According to the idea of purpose and capabilities of the system, it must cope with the task.

 
mytarmailS:

because I have the data to be called "d" ))

insert your

clos <- dt$your closing price

Heh, what a hassle with that R

пакет '‘uwot’' недоступен (for R version 4.0.2)

В 3.5

Предупреждение в install.packages("TTR", "uwot") :
  'lib = "uwot"' не для записи

what does it mean?

 
Aleksey Vyazmikin:

Heh, what a hassle that R

В 3.5

What does that mean?

Try this.

install.packages(  с("TTR","uwot")  )

I messed up

 
mytarmailS:

Try it this way

I messed up.

Ошибка в с("TTR", "uwot") :не могу найти функцию "с"
 
Aleksey Vyazmikin:
Maybe the "s" in Russian was typed?
 
elibrarius:
maybe you typed the "s" in Russian?

I copied :)

 
elibrarius:
maybe you typed the "s" in Russian?

It downloaded, it was indeed written in Russian.

> library(uwot)
Загрузка требуемого пакета: Matrix

> train.idx <- 100:8000

> test.idx <- 8001:10000

> UM <- umap(X = X[train.idx,],
+            y = Y[train.idx], 
+            approx_pow = TRUE, 
+            n_components = 3, 
+            ret_mode .... [TRUNCATED] 
Error in x2set(Xsub, n_neighbors, metric, nn_method = nn_sub, n_trees,  : 
  Non-finite entries in the input matrix

But some other package is needed...

Can't he himself think that it should be downloaded...

 

not an advertisement, but if you pay attention to one bot (several) in the top in the market, you will see how certain watches are traded )

curiously, the bot is published by date later than my article (perhaps the author was inspired), but I did not get that cool results. Apparently, I wasn't really looking there.

The tree forests aren't so cool either, although I've seen some nice variants.

***

 
Aleksey Vyazmikin:

It downloaded, indeed it was written in Russian.

There's another package I need...

Can't he figure out that he has to download it himself...

It doesn't need any package, where does it say that?

Non-finiteentries in the input mat rix >>> Non-finite entries in the input matrix

>> maybe there is an inf.

Maybe you should try to understand some elementary syntax and only after that train your models?