Machine learning in trading: theory, models, practice and algo-trading - page 798
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
He meant that he already has the MO he needs, it's just a matter of a small, but huge thing for us
Which is it? don't be lazy to explain.... :-)
Which is it? don't be lazy to explain.... :-)
Here it is:
Forum on trading, automated trading systems and testing trading strategies
For more information about the theory and practice of machine learning in trading (trading and not only)
Maxim Dmitrievsky, 2018.03.31 14:27
i am interested in analysis of variance and everything to do with probabilities (including conditional ones) and non-stationary series (new methods, developments)
i have already finished studying mO and took what i needed from it
If I have something to say on the subject I will discuss it, all the rest seems to me nonsense and a step too far.
I see. So you are no longer interested in the MO... Sorry I didn't know....
I've written many times before that the methods used here are overwhelmingly unsuitable for time series forecasting
new studies are increasingly touching on this topic, but it is not discussed here
the only thing that someone suggested is Alexander, but I think that one should look somewhere in the neighborhood, not there
the probabilistic approach was also mentioned here by Yury Asaulenko, but neither he nor anyone else could show or reveal the potential of the topic
I am now writing about non-parametric methods where only price and its derivatives act as a predictor
Personally for me it's always been a goal to find an algorithm of actions, which would always lead me to a consistently good result, despite its volatility. It's a bit like On. Off. Money. Isn't it beautiful. Always do the same actions and get the same good result and don't bother, except from boredom and curiosity for... :-)
Well, that's the way it should be. It's "off, dough" and nothing else.
)
I have written many times that the methods used here are overwhelmingly unsuitable for time series forecasting
new studies are increasingly touching on this topic, but it's not discussed here
the only thing that someone suggested is Alexander, but I think it's necessary to look somewhere in the neighborhood, not there
the probabilistic approach was also mentioned here by Yury Asaulenko, but neither he nor anyone else could show or reveal the potential of the topic
I am now writing about non-parametric methods, where only price and its derivatives act as a predictor
What about me Max? I solved the problem and got a stable method of obtaining models that are guaranteed to work in the future, even if not long as I would like, but enough to really earn......
What about me Max? I solved the problem and got a stable method of obtaining models that are guaranteed to work in the future, albeit not long as I would like, but enough to really earn......
what are you talking about all the time? your system is zero on average
What are you talking about all the time? You have a system at zero on average.
Well, you are now witnessing practical proof of the adequacy of the approach. I am sure that you will also have it when you move from the field of research to practice. As a rule at this moment you already do not search and do not experiment, and simply use result of your persistent labor for many years. IMHO.....
It is clear that the study can not stop and move on and evolve, but it is purely out of boredom, because the robot rules and it is better not to interfere, and my hands itch because I'm used to look for something and invent different algorithms, not forgetting to maintain the basic TS in working order, spending on it as much time as it requires.
For example, I'm now going to work on soccer betting, which will be another and certainly significant proof of an adequate understanding of machine learning. If the result is positive, of course...
Because getting equally good results in completely different areas is exactly what is called professionalism.
Suppose you have a system that performed well in forensics and then you get an offer to use it in medicine. What? Are you going to refuse?
Well, you are now witnessing practical proof of the adequacy of the approach. I am sure that you will also have it when you move from the field of research to practice. As a rule at this moment you already do not search and do not experiment, and simply use result of your persistent work for many years. IMHO.....
It is clear that the study can not stop and move on and evolve, but it is purely out of boredom, because the robot rules and it is better not to interfere, and my hands itch because I'm used to look for something and invent different algorithms, not forgetting to maintain the basic TS in working order, spending on it as much time as it requires.
For example, I'm now going to work on soccer betting, which will be another and certainly significant proof of an adequate understanding of machine learning. If the result is positive, of course...
Because getting equally good results in completely different areas is exactly what is called professionalism.
Suppose you have a system that performed well in forensics and then you get an offer to use it in medicine. What? You're gonna say no?
What are you, nuts? Where's the proof?
you shouldn't have asked, i'm sure "at all".
Are you nuts? Where's the proof?
You shouldn't have asked. I'm sure it's "in general."
What kind of proof do you need? I don't understand... Isn't real time trading the most reliable proof. Or do you still want to see everything on the history in the tester....