Machine learning in trading: theory, models, practice and algo-trading - page 3427

[Aleksey Vyazmikin]

mytarmailS #:

I criticised your attitude to risk, but I didn't gloat anywhere.

I see. I hope I explained my attitude to risks and why there are no stops. Stops are needed when there is a signal to change the trend.

[mytarmailS]

Maxim Dmitrievsky #:

Been into this for a long time. There are books:

Non-Fractal Markets. Mandelbrot.

Fractal theory of the forex market. Almazov.

In general, the analogy is traceable.

like this

weierstrass <- function(x, a = 0.5, b = 3/2, n = 100) {
    sapply(x, \(x) sum(a^n * cos(b^n * pi * x)))
  }
  
x <- seq(-2, 2, by = 0.01)
y <- weierstrass(x) |> cumsum()
  
plot(x, y, type = "l")

[Maxim Dmitrievsky]

Aleksey Vyazmikin #:

I don't want just a tree, I want a sequential clustering. Like in the diagram below. The number of clusters at each level should be set arbitrarily, as well as the depth of the tree.

https://scikit-learn.org/stable/auto_examples/cluster/plot_agglomerative_dendrogram.html#sphx-glr-auto-examples-cluster-plot-agglomerative-dendrogram-py

[Maxim Dmitrievsky]

mytarmailS #:

such as this

Yeah.

here's a paper from the '80s.

https://royalsocietypublishing.org/doi/epdf/10.1098/rspa.1980.0044

[mytarmailS]

Maxim Dmitrievsky #:

Yeah

here's a paper from the '80s.

https://royalsocietypublishing.org/doi/epdf/10.1098/rspa.1980.0044

It's all alchemy to me.

[Maxim Dmitrievsky]

mytarmailS #:

It's all alchemy to me.

groovy

[Aleksey Vyazmikin]

Maxim Dmitrievsky #:

https://scikit-learn.org/stable/auto_examples/cluster/plot_agglomerative_dendrogram.html#sphx-glr-auto-examples-cluster-plot-agglomerative-dendrogram-py

Why this link - it seems to have already been discussed and worked out why it doesn't fit...

[Maxim Dmitrievsky]

Aleksey Vyazmikin #:

Why the link - I thought it was already discussed and sorted out why it doesn't fit....

I haven't really looked into it, but it might be useful.

[Aleksey Vyazmikin]

Maxim Dmitrievsky #:

I haven't really looked into it, but I just thought it might come in handy.

Thank you.

[mytarmailS]

Maxim Dmitrievsky #:

groovy

weierstrass <- function(x, a = 0.5, b = 3/2, n = 100) {
    sapply(x, \(x) sum(a^n * cos(b^n * pi * x)))
  }
  
  seq(0,10,length.out=n) |> 
    weierstrass() |> 
    cumsum() |> 
    xts(Sys.time()+1:n) |> 
    to.minutes5() |> 
    chart_Series()
   

just like the market ))))))