Machine learning in trading: theory, models, practice and algo-trading - page 3196
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Well including )
This is a continuation of the topic
Forum on trading, automated trading systems and testing trading strategies
Machine learning in trading: theory, models, practice and algo-trading
Aleksey Nikolayev, 2023.08.17:45 PM
I can suggest modifying my experiment. Let there are ten boxes with numbers from 1 to 10, one hundred white balls and one hundred black balls (numbers 10 and 100 are taken conventionally). The balls are somehow arranged in the boxes, then you look how many balls are in each box and try to understand if there is a regularity in the algorithm of arrangement - in the boxes with which numbers there is a predominance of balls of some colour.
So, if each ball (of both colours) is just put randomly and with equal probability 0.1 in one of the drawers, then in the end there will be no uniformity in the ratio of colours! Almost always there will be a box where almost all white and one where almost all black. And the matter is not at all in quality of DSP, you can take a real quantum DSP and everything will be the same. It's about the very nature of probabilistic randomness. There will always be irregularity, but the numbers of boxes where it will be found at the next layout are absolutely unpredictable. It is the same in the previous example with the hour of the week (the hour of the week is the analogue of the box number).
There are two ways to do this. Either try to show that the unevenness in practice is much greater than it would be under equal probability. This is done by some kind of statistical tests. Or just be sure that the non-uniformity, though small, is due to some regularity, which is just weakly manifested due to noise. But that's a matter of faith and practice and if it works, ok.
I hope it's clear that the box numbers (hour of the week) are an analogy to your quanta.
It's a continuation of a theme
Why do you need to change the targets?
you need to take many variants of the original shuffled series and calculate new targets for them.
then run 10k simulations and look at the best average quartiles and compare them with the original ones.
SanSanych, is there any chance that your mt-R package will work in Linux? I mean both possible variants of R installation - directly in linux and via wine (I haven't tried this variant yet).
The reason of interest is that they are going to close windows for Russia.
SanSanych, is there any chance that your mt-R package will work in Linux? I mean both possible variants of R installation - directly in linux and via wine (I haven't tried this variant yet).
The reason of interest is that they are going to close windows for Russia.
What are the advantages of this package over other options?
and why change the target ones
you need to take many variants of the original shuffled series and calculate new targets for them.
then run 10k simulations and look at the average better qua segments and compare them with the original ones.
You've lost the context - that's not what you were talking about.
If about your idea, where do you get 10k from - do you want that many time series? That's too much - not rational.
There is another option - take other tools and try there to see what percentage of quantised segments continue to be effective.
This will highlight common patterns on different instruments.
You've lost the context - it was about something else.
If it's about your idea, where do you get 10k from - do you want so many time series? It's too much - it's not rational.
There is another option - take other tools and try there to see what percentage of quantum segments continue to be effective.
This will highlight common patterns on different instruments.
Why different instruments? You can do it on one.
you can do less, it makes no difference, but not too little (the strength of smallness is determined intuitively and empirically).
the more simulations, the more averaged the result. Bias is a variant of tradoff.
That is, when reaching some threshold of the number of simulations, quatraffic will cease to be useful in terms of profit extraction, will be too averaged. That is, they will get approximately equal importance. This is if the market is really random. If you get to that stage, you will actually consider the market to be a random. Well, or your chip-value bundles are inefficient.
will be a normal monte carlo of a healthy person )
SanSanych, is there any chance that your mt-R package will work in Linux? I mean both possible variants of R installation - directly in linux and via wine (I haven't tried this variant yet).
The reason of interest is that they are going to close windows for Russia.
Looked through the news - I see only one year old restrictions on downloading W10 and W11 distributions and their removal at the end of 2022. Is there an update somewhere?
I don't need any others, you can use one
you can have less, it doesn't matter, but not too little (the strength of smallness is determined intuitively and empirically).
the more simulations, the more averaged the result. Bias is a variant of traidoff.
That is, when a certain threshold of simulations is reached, the quattrograms will cease to be useful in terms of profit extraction, they will be too averaged. This is if the market is truly random. If you get to that stage, you will actually consider the market to be random. Well, or your chip-signs bundles are inefficient.
will be a normal monte carlo of a healthy person )
As I understand it, you propose the following in essence:
Maybe I could do a couple of dozen samples, but:
1. I have no understanding of how to generate a truly similar symbol to the original - no tool.
2. How do you propose to interpret the result if there are many "hits" and, if few, in the ranges of quantum segments?
That package runs on dlls. And they only work in Windows.
Looked through the news - I see only one year old restrictions on downloading W10 and W11 distributions. And about their removal at the end of 2022. Is there an update somewhere?
They said that activation will be switched off by key for corporate customers.
As I understand it, you are proposing the following in essence:
Maybe I could make a couple of dozens of samples, but:
1. I have no understanding of how to generate a sample really similar to the original - no tool.
2. How do you propose to interpret the result if there will be many "hits" and, if few, into ranges of quantum cutoffs?
Randomly shuffle the symbol n times, look at the average of the quantised cutoffs (which are the best). If on the average none are the best, then there is nothing to lose there.