Machine learning in trading: theory, models, practice and algo-trading - page 2467

[Evgeny Dyuka]

elibrarius #:
A picture of 3-5 deals tells you nothing. 100-200 real trades made (or in 3-4 months) will show the statistics. And the drawdowns.

This forum is already dead because it is affected by slackers and lazybones.

[Forester]

Evgeny Dyuka #:
This forum is already dead because it is affected by idlers and lazy people.

Yes. Everybody is too lazy to trade on their own indicator for 3 months)))

[Valeriy Yastremskiy]

In general, it seems that the rule, No one owes anyone anything, does not work here))). It's a pity. Communication would have been more constructive.

[mytarmailS]

Interesting service for those who are interested in "trading against the crowd"

https://fxssi.com/beginners-guide

[transcendreamer]

Evgeny Dyuka #:
This forum is already dead because it is affected by idlers and lazybones.

The fact is that all have long been sitting in telegram sects, there all the movement and secret knowledge, and the forums unfortunately are almost dying out 😕

[mytarmailS]

Does anyone know how to parse graphs like this?

[Dmytryi Nazarchuk]

mytarmailS #:

Does anyone know how to parse charts like this?

You do realize that forex prices are not driven by clients of these "brokers", right?

The sample is not representative.

[mytarmailS]

Dmytryi Nazarchuk #:

You do realize that it's not the clients of these "brokers" who move Forex prices, right?

The sample is not representative

I understand it very well.

[Valeriy Yastremskiy]

transcendreamer #:

The thing is that everyone has long been sitting in telegram sects, there is all the movement and secret knowledge, and the forums, unfortunately, almost dying out 😕

I'm waiting for knowledge), but the mountain will not come to Mohammed. By the way, the usability on the trolley is worse than here.

[JeeyCi]

Maxim Dmitrievsky #:
Here we have such a scheme, or rather, there can be two of them.

1. First the search area is determined: empirically or on the basis of assumptions, statistical tests are made. Then an MO algorithm is chosen which fits better. Then the model is meaningful.

Yes, I agree with you - we don't come up with probability distribution from our head - it's clear that in the long run many/some (Poisson, binomial and other symmetric distributions) lead to normal - but this is the lack of trading opportunities (when probabilities are distributed normally)... the correct way: test the null hypothesis of a particular/any distribution for plausibility (by comparing the available distribution and the theoretical/standard distribution), if the statistic fails, then accept the alternative hypothesis... you'll just get buried (me so far, if I start) to determine Which distribution (e.g., option prices) if you rely on statistics, and you can't search for probabilities without it...

I found it by the way(what I tried to remember 2 pages ago - when no one refuted it - but thank you for the reminder):

Still, price distribution is normal, and profit log-normal distribution has(i.e. a lot of probability for small losses and little probability for large profits) - in general, asymmetrical - I tried to remember... I tried to remember... In general the Black Sholes formula has a binomial distribution, but they say it doesn't work... I have to look for the distribution, as it is scientifically defined in the stat. Processing - a bit long (to prove all null hypotheses, until you get either Poisson's, or log-normal, or some other specific distribution with proven reliability of choice by Fisher's criterion, at least, to finally start from this distribution for interpretation of probabilities)...

so I'm still cautiously looking at statistics and theoretical beliefs - though I'm interested in SKEW CBOE- there's a link to Cboe SKEW Index Whitepaper - my calculations are still a bit floundering... (

But I want to see the term-structure in perspective (although this SKEW is calculated for the S&P and spike-up says about the fall, but I want to see how it will behave for the currency)... imho (at least on the history)