The essence of optimisation - page 12

 

one day in a science class, the teacher asked what you saw on the blackboard (the blackboard was clean)... the students described what they saw. the teacher said "we listened to 10 opinions and they all differed in some way, what are the differences in perception?

what is the reason for the difference in perception? - in the individuality of vision, in the number and quality of visual receptors? - Maybe because of the different way in which visual images are processed? - if there are so many opinions on such a simple "truth" as a blank blackboard, then maybe there is not really objective and reliable available information on the subject? maybe the world around us is illusory?

I mean, it seems ridiculous to hear those who talk about "crowd psychology" and other such nonsense... what kind of crowd psychology is it if ten people cannot see a clean blackboard the same way...?

 
VNIK:

On point 1 --- everything is correct and logical, except for the last phrase: "trying to predict how most will act in the near future" - this is the author's own speculation

But as for point 2. --- There is no logical connection between p.1 and p.2, as p.1 is pure mathematics and p.2 is pure metaphysics.

All in all, on the subject of logic, there are tensions here.

Mathematics is a language, axioms and rules of inference, nothing more. The magic of mathematics lies only in STRICTLY following the rules of OWN axioms and the logic of generating new symbolic sequences from axiomatic sequences. Actually, any programming language is an extension of mathematics, because everything remains the same, but most people, especially those who did not study in those universities and basics of cybernetics, strongly emphasize "mathematics" especially its old types (like analysis) as something mystically super correct, which can create something new, which is perceived as a miracle by ignorant mathematicians, as the priests' tricks were perceived by plebs before.

I think if the argument stumbles on the fact that programming is not part of mathematical description, we can stop talking, because it is not good to cast aspersions as they say ...

Otherwise there would not be a question that mathematics in the right hands can lie, as well as political demagogy in the mouth of a politician, all depends on the skill.

But the interesting thing is that maths has nothing to do with it)))) "Point 1" is precisely pure psychology, a kind of psychology, several types of psychology, "social", "cognitive", "behavioural", etc. etc. Empirical sciences of the inductive type. By the way, it is possible that those who do not know that the origins of correlation and factor analysis are in psychology, psychologists were the first to invent the quantitative identification of common "factors", "principal components", etc. Ask yourself.

Statistics were invented by psychologists in short. Some people still don't even take it for a science.

joo:

once, during a science class, my teacher asked me what I saw on the blackboard (the blackboard was clean)... the students described what they saw. the teacher said, "we have listened to 10 opinions, and they all differed in some way. what are the differences in perception?

what is the reason for the difference in perception? - in the individuality of vision, in the number and quality of visual receptors? - Maybe because of the different way in which visual images are processed? - if there are so many opinions on such a simple "truth" as a blank blackboard, then maybe there really is no objective and reliable available information on the subject? maybe the world around us is illusory?

What kind of crowd psychology is it, if ten people can't see a clean blackboard in the same way...?

You are talking about the old dilemma of model and reality, which is essentially another interpretation of the "main philosophical question" of subjective and objective.

There is no unambiguous, consistent logical answer capable of tilting to either side when the debaters are of equal intelligence.

The answer lies in the utilitarian usefulness of a model and its explanatory qualities, which are also subject to utilitarian tests.

In general, if a model is beneficial it is correct.

In Popera's interpretation it must be "falsifiable", but with respect to models of high-level "outgoing" processes generated by people, a mere benefit with respect to survival and replication by the marathon participants using it is sufficient.

 
Alex_Bondar:

Lots of words.

They'd better think of a way to attach the Ataman's indicator to the whole thing.

 
Alex_Bondar:

Otherwise there would be no question that mathematics in capable hands can lie, just as political demagogy in the mouth of a politician, it all depends on skill.


Perhaps (by a wide margin) you are right. But what prevents you from changing your phrase by substituting "politician" ---> "psychologist" and we get it:

Otherwise there would be no question that mathematics in capable hands can lie as well as psycho demagogy in the mouth of a psychologist, it all depends on skill.

And where do we get to?

But the thing is that mathematical calculations can ALWAYS be checked and verified, while the psychologist's statements have to be taken on faith...

 
VNIK:

But the point is that mathematical calculations can ALWAYS be checked and verified, whereas the psychologist's claims have to be taken on faith...

+

 
Alex_Bondar:

Yesterday I finally finished one interesting algorithm... I won't even tell you what it's about, it's so cool.

Auto-optimization is still a cool thing, but it requires a lot of effort and a lot of innovation.

No, no... No good... share your impressions, enthusiastic shouts... what inspired this epiphany, where do the legs of the mega-grale come from? Let's rejoice together over your success.
 
toxic:


For example strategy: Random entry, TP and SL locking.


I've highlighted in bold what removes the whole point of optimisation.

Sorry - didn't mean to necro-post...

 
The only purpose of optimization - is to determine: On average, on what number of historical data (the optimization period) the positive qualities of the theoretical calculations, justifications of inputs and outputs, logic of its behavior and other, independent of the optimization, ways to achieve the goal, on condition TP=SL = 1000 or more points of 4-digit (protective stop-order) are best of all available in the TS. Moreover, such optimization should be conducted using all available instrument history, without dividing it into "Back" and "Forward" areas. There are two possible modes of setting up the TS - trend and flat. We need to accept the fact that the trend TS will die in the flat, and the flattened one - in the trend. The intermediate variant of TS, working both with a trend and with a flat, can be developed, but its effectiveness will be low. If there is no theoretical substantiation of the TS logic, the optimization will not help, moreover, it will prove to be harmful, leading the trader along a false path. This is purely my personal opinion.