Machine learning in trading: theory, models, practice and algo-trading - page 2328

[Valeriy Yastremskiy]

Maxim Dmitrievsky:

If you discuss incomprehensible things for ten pages and complain about misunderstanding, there will be no understanding.

I am obsessed with a pink book that I can't finish reading. The stone flower cannot be solved and put together)).

And substantively, except for the results of using packages, we can not discuss. The structure of a grid is not clear in some places. And the search area is price and time. There is no point in discussing them.)

[mytarmailS]

Vladimir Suslov:

Resistor and two diodes = Profit
)

Yeah, I'll take a soldering iron and make a robot for forex )

But seriously, don't you see the difference between my example with the market and your example with the sine wave? ??

Maxim Dmitrievsky:
I give you a lot more wisdom than in statistics books.)

Ahhh, he loves himself so much.

Top of the line wisdom in MO is to do averaging, renata would approve it 100%.

Oh yes, the zen is done!

[Maxim Dmitrievsky]

mytarmailS:

Yeah, I'm going to take a soldering iron and make a forex robot.)

But seriously, don't you see the difference between my example with the market and your example with the sine wave? ??

Ahhh, how much he loves himself it's fucked up...

And the tip of the wisdom in MO is to do an averaging, renata would approve 100%

Oh yes, the zen is learned!

When you don't know what to say, you better sing

[mytarmailS]

Does anyone have a theory as to why these "market profiles" work?

function to build these profiles directly on the chart, and not just on prices

loc.aria <- function(X,y=NULL,breaks=21,cont=0){
loc <- locator(n=2)
id <- round(loc$x[1]:loc$x[2],0)
x <- X[id]
h <- hist(x,breaks = seq(min(x,na.rm = T),max(x,na.rm = T),length.out = 21),plot = F)
prc <- h$mids
cnt <- h$counts
if(!is.null(y)){
hy <- hist(y,breaks = seq(min(y,na.rm = T),max(y,na.rm = T),length.out = breaks),plot = F)
cnt <- hy$counts/cont}
col <- rgb(0,0,1,alpha=0.2)
segments(id[1] , prc , id[1]+(cnt*10) , prc ,col = col,lwd=5)
abline(v=range(id),col=8,lty=3)}

[Mikhail Mishanin]

mytarmailS:

Does anyone have a theory as to why these "market profiles" work?

function to build these profiles directly on the chart, and not only on prices

I guess - the human factor) Everything that is deposited in the under/over etc. mind of humans somehow works)

[mytarmailS]

Mikhail Mishanin:

I think - the human factor) All that is deposited in the sub / above and so on the minds of men somehow works).

And if I say that these are not prices but a normally distributed random))

[Farkhat Guzairov]

mytarmailS:

And if I say that these are not prices, but a normally distributed random ))

Then this kind of "Market Profile" really isn't. )))

[Mikhail Mishanin]

mytarmailS:

And if I say that it's not prices but a normally distributed random ))

It makes no difference) the only question is how many people will regularly see it.

[Rorschach]

numerical methods, still

[mytarmailS]

Maxim Dmitrievsky:

Did you close any trades with your hands? Why do you have Algotrading: 96% and not 100%?