Machine learning in trading: theory, models, practice and algo-trading - page 1822

 
tamagucci:

On the slipper terrors (Glock) on the weekend try) With birst. I have not played for 7 years, but the skill is not lost)

P.S. You can just on the guns. The most fun. Deagle does not count ;)

Practise while there.

That's it... I've started training. My nickname is Gramazeka if you know what ....
 
elibrarius:
Do you mean that price quotes are almost random straying?
It turns out that so
 
elibrarius:
I read the first few search results for multi stage sampling. There's just about the cost savings in doing research, not to survey 300 million people, but 1,000 by randomly choosing 10 states, 10 districts, 10 people in them, and only surveying them.
We have all the quotes for many years in our tasks. There is no cost to get them.
Apparently you found something else of interest? Not what is in the first search results.
Nothing particularly interesting, you can just add sampling by condition. More often or less often, depending on the condition. In general, random sampling should be better.
 
ipsec:
Hi Maxin,

Eu put only six features of last 3 ticks
MA1(SMO) and MA5(SMO) with standardization.
With these features I put to train only if price of last MA1 > bollinger bands upper or MA1 < bollinger bands lower to reduce my time of training.
After training one day in 2018 (take 24hours of training to get lower epsilon) my model is ok to back testing.
I tested in Jan 2019 and getting good results.

So I think the number is dimensions and features is not very relevant if the feature is good enough.

Best regards,
Fernando
Well, I see. But I can achieve same results after 1-5 min of learning :) if features are good
 


Do you mean that price quotes are almost random ramblings?

Maxim Dmitrievsky:
It turns out that

Since the market consists of transactions it turns out that :

We make decisions by chance?

We open trades by chance?

place stops accidentally?

optimize the system accidentally?

economies function by chance?

 
Uladzimir Izerski:

A pattern is a visual-mathematical model that can be identified in advance. I haven't found any other way to see the future in many years .

I.e., its pattern can be predicted in advance, and what else is needed?

So the pattern is already over when it is recognized, right? That is, I want the entry direction to be repeated in the same direction...

 
mytarmailS:

Since the market consists of transactions, it turns out that :

we make decisions by chance ?

open trades by chance?

place stops by chance?

we optimize systems by chance ?

economies function by chance?

what kind of philosophical questions are you asking ?)

It depends on what you're relying on. If on Vedanta, everything is an illusion. If on the big bang theory, then everything is an accident. Matter happens to outnumber antimatter by 1% and now you're sitting here.

And if everything is an illusion, then it's unknowable, because there really isn't any, there's nothing to know

 
Maxim Dmitrievsky:

What a philosophical question you're asking.)

It depends on what you're going to base it on. If it's Vedanta, then everything is an illusion. If the big bang theory, then everything is an accident. Matter happens to outnumber antimatter by 1% and now you're sitting here.

The questions aren't philosophical, they're rhetorical...

It's clear that it's not random...

hence the conclusion that the wrong tools on the market experiment, hence the wrong conclusions...


If I take a hypothetical risk for a second and imagine that only the support and resistance levels work in the market, it becomes clear at once that working with returns, working in the sliding window, to put it mildly, makes no sense.

 
mytarmailS:

The questions are not philosophical, but rhetorical...

It is clear that it is not by chance...

hence the conclusion that the wrong instruments for the market experiment, hence the wrong conclusions...


If we take a second, hypothetically, and imagine that only the support and resistance levels work in the market, it becomes clear at once that working with returns, working in the sliding window, to put it mildly, makes no sense.

who understands? ))

 
Maxim Dmitrievsky:

who understands? ))

To everyone who thinks, including me.