entrance point - page 10
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
In systems theory there is a postulate (I do not remember exactly how it sounds, but I will give the meaning): the performance of a system is greater than the sum of the performance of its elements separately...
I think it was about the probability of failure of a complex system.
So, the reliability of a system is always less than the reliability of its most unreliable component.
I think it was about the probability of failure of a complex system.
Well, the reliability of a system is always less than the reliability of its most unreliable component.
>> no, no... not about reliability... I don't think 'performance' was mentioned there either (that's how I replaced it)... I think it was about the energy of the system... I remember that I also made a conclusion: an organization of people (according to this postulate) is always stronger than all of them individually...
no-no... not about reliability... I don't think 'performance' was mentioned there either (that's how I replaced it)... I think it was about the energy of the system... I remember that I made a conclusion: an organization of people is always stronger than each of them separately...
There is no point in talking like that :) . >>) : I need a precise formulation to continue the debate.
There is no point in talking like this :) . >> .
Unfortunately, I don't have the book where I read it: "Production Organization" by R. A. Fatkhutdinov. I didn't find it on the Internet... I can't read other books, but if I come across this postulate I'll let you know :)
I think it was talking about the probability of failure of a complex system.
Well, the reliability of a system is always less than the reliability of its most unreliable component.
that's absolutely right reliability is less but not performance
but then appears a publication stating that Japanese have come to conclusion that increasing parameters of an economic model leads to deterioration of its behavior
and the optimal number of parameters is only 7-8, not hundreds or thousands
It seems clear,
But here we have a TS and what should be considered its parameters?)
If each parameter of the TS were an indicator, then MTS with 7 or 8 indications would be viable
But it's not so, the MTS with 7 indicators would give very good results in the optimizer (it's clear why)
and will turn out to be completely unworkable outside the optimizer
== in particular - one of the issues is to calculate the linear dependence of the indicators used.
In systems theory there is a postulate (I don't remember exactly how it sounds, but I'll give the meaning): the performance of a system is greater than the sum of the performance of its elements individually...
the negative aspects may lie in two things:
- in a character of connections of these elements (experts in our words)
- in the number of elements (too many EAs will decrease, not increase "productivity")
sorry for being so abstract... I myself have all these ideas at the stage of development only...
the link is not sequential but partial. the advisors do not depend on each other, their work results are taken as a proprietary value.
the more they work, the better: all strategies and even the fundamentals of the system, the total efficiency (if the work is sequential, of course, it would decrease) what we need is a special case of the possibility of their convergence, which can then be expressed in absolute or relative values
if every parameter in TS was an indicator, then MTS with 7 or 8 indices would be viable
but it's not so, MTS with 7 indicators would give very good results in the optimizer (it's clear why)
and will turn out to be completely unworkable outside the optimizer
== in particular - one of the issues is the calculation of linear dependence of the indicators used.
I agree... as Better said: fitting can be right and wrong... in my opinion too - too many parameters leads to poor results on OOS... but you have to keep in mind, Korey, that one indicator usually has more than one parameter... that's why usually 2-3 indicators give up to 7-8 parameters for optimization...
the more the better with this system
typical beginner's point of view... it's not that simple...
Finally got to the computer. :) With a parallel link, you'll get a drain. You need to create a serial link. In simpler words. What's the best way to purify water? By running it through 3 filters simultaneously or by putting them in series one after the other. Which is the better product?