Where is the line between fitting and actual patterns? - page 39
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
Absolutely. But nobody will forbid us to reoptimize (in a good sense) the TS at the end of each pattern, i.e. to check whether the patterns have changed. Thus, the buffer of useful patterns can always be kept larger than the volume of changing patterns. The advantage of the second type over the first is the ability to observe and track changes in identified patterns. In the case of the first type it is not even possible to reveal them.
Read what you are told more carefully. Check if it's there.
Second question - how do you do it without a history?
Got it.
Sorry, still thinking about the first one.
Well TS of the second type do not guarantee profitability in the future, as well as TS of the first type can be profitable. The second type was introduced in order to have clear criteria and parameters of optimization and to be able to identify whether an TS is fitted or optimized, strictly speaking, to separate flies from shameless cutlets. Ugh, on the contrary. :)
Probably not a bad option here for type 2, truth is not in the code, but the potential, I'm sure, is there. Myself still "approaching" a dig in this direction, although the topic and
old, but the approach is the same. I read somewhere that there is a real earning expert who exploits this pattern, which is why I identified
for my consideration - http://forex.kbpauk.ru/showflat.php/Cat/0/Number/25750/an/0/page/0#Post25750.
Where is the line between fitting and real patterns?
Looking at the market we see that possibly existing patterns cannot be parametrically constant. Every system has a level of fit and a level of regularity of one or more events.
And the preponderance towards the second level is responsible for the rationality of the trading idea itself.
Thinking abstractly. The thoughts of others will be of interest.
I do tests on minutes with a period of 5 years. No tweaking helps in any strategy I have tested. I would like to know how others are doing?
I have a sample period of max. 4 months, of course it depends on the system to what extent it "holds" the patterns from the sample in the future.
questions:
1) why do you need backOOS? extra test, has it given anyone any good?
2) marshmallows and other mardas in the form of crocodiles and similar crawling creatures ) are the worst examples of TS!
2)Mashkins, macdashkins and other mardashkins in the form of crocodiles and similar crawling creatures .....
Oh, my God! Mashka is such a creature. It is obvious that she is a crocodile:))))
1)what is backOOS for? an unnecessary test, has it done anyone any good?
backOOS and forwardOOS are the same thing.
Let me explain, if we are not talking about "eternal" patterns that are always in the market, but about coming and going, then the pattern appears at some point and leaves at some point. The probability that our Sample period will coincide with the moment the pattern appears is approximately equal to the coincidence of its end with the disappearance of the pattern. By conducting OOS after Sample we increase the chance of not catching a pattern in the market. By conducting it before Sample, we bypass this rake.
In fact, this is a very primitive judgement, a pattern does not appear overnight and does not disappear at one moment. There are always regularities in the market. There are many of them. Perhaps it looks like a pile of sinusoids superimposed on each other with a certain pitch. What is worse, there are a lot of these piles with various periods. Selecting the Sample size we select we select the period of this sine wave. After we run Sample, we should catch the sine wave falling maximally on the Sample time section (this pattern is the easiest to spot). There should be tails in front and behind. What is in the front may be traded, the one in the back may be used for testing.
I use only backOOS for my TS, fOOS can be used for tests and experiments. In principle this approach justifies itself for me.
Z.U. I wasn't able to express my thoughts clearly, but honestly, I tried)
Who considers the Breakeven subsystem to be an integral part of TS?
What is it?
I am sorry, but you did not understand the question - in such association of sets it will be impossible to identify regularities by known methods - "figures" will be different not only in size - they will be absolutely incomparable between each other, i.e. you will certainly find similar figures, but they will be similar only in form, but not in content.
I apologise, but I did not understand the point of the answer.
A figure in terms of content is what?
Usually, the shape should correspond to its content (the established behaviour of the process in the future).
And what is this substitution - look alike!, and the result is again 25...
% of guessing apparently.
;)