Probability, how do you turn it into a pattern ...? - page 90

 
It's too late to drink Borjomi when your liver is already damaged...
 
Andrei01 писал(а) >>
Doesn't look like a Neuveteran to confuse cents with dollars.


Yes, all the more so Nevteran is like Karabas the Barabbas - he exploits small children (just kidding :))
 
Neveteran писал(а) >>

Dimitri, what is the point of demonstrating something that is highly questionable?

 
here I read http://www.liveinternet.ru/users/rusotechestvo/post118457050/ and I was struck by the phrase:
"Decision-making procedures take into account their expected utility, which is expressed as the product of the expected income by the probability of receiving it."
 
moskitman >>:
вот прочел http://www.liveinternet.ru/users/rusotechestvo/post118457050/ и меня поразила фраза:
"В процедурах принятия решений учитывается их ожидаемая полезность, которая выражается как произведение ожидаемого дохода на вероятность его получения."

Discovered the mate expectation? Good! Better late than never.

 
from Wikipedia: The mathematical expectation is a measure of the mean value of a random variable in probability theory.
 
moskitman, Wiki is not serious. A formula is needed here.
 
moskitman >>:
из Википедии: Математическое ожидание — мера среднего значения случайной величины в теории вероятностей.

Right. Making a profit in any venture of little to no risk is a random process (just like the time the bus arrives at the bus stop, how many years you will live, winning the lottery, and even getting your paycheck), and the "expected utility" is the average income, which is calculated according to the formula mentioned above. The same formula is in wikipedia

In this case x(i) the expected income, p(i) the probability of getting that income.
In matters of mathematics and probability theory, wikipedia is quite authoritative. Especially the English version, which is much more complete than the Russian version.


 
timbo >>:

...............

В вопросах математики и теории вероятности википедия вполне авторитетна. Особенно англоязычный вариант, который намного полнее русской версии.


By the way, yes - I mean the English articles. The material is fuller and, as a rule, better presented in a different way. In the Russian version, sometimes there's so much contextless stuff that you can't understand what they're talking about.

===

And sometimes there are no topics in the Russian version, only in English.

 
Literally "a few words" ... No pretensions, sort of a question for those present.

Yesterday I got into a fascinating conversation in a rather reputable DC. (live)
There was a discussion about TA methods for the right to exist. There were about 10 people from each side, elioticians, fractalists and networkers. After long talks about the coolness of the approach, came to the conclusion that ZUP and some ZZ with channels, in H4 give 70% of correct entries, it seems to be agreed upon by netizens. Then we started discussing stops, trall, and short profit closing. And everyone unequivocally rejected working without stops and of course martin with a margin. Then comes the focus and people go back to technical peculiarities of stop and profit ranges for entering on the signal from the indicator. And so we have decided referring to the best case study of one of this forum participants (SL_to_Bar), that a 4 times larger profit at the primitive level is a good example for correct and quality processing of 70% of correctly predicted market entries with the technical indicator. Then came the collective euphoria of superior self-awareness and I left.

Later I figured, in a very simplified version, further developments:

if conditionally profit = 10 points and stop 40 ka, then 7/3 gives +70/-120, what a great idea!!!
variant with lots of false entries (stop lossless closing) in search of trall, will not give the best results...

I would use signals of indices (portfolio) without stops and profits, but I would close profit or loss on balance equi within specified limits (for example, the amount of pledge in + or -) although it's not so cool.

Then, what is the point of non-probability inputs????