Now let's be clear, do patterns work? Let's discuss) - page 4
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
We should punish him. Put him in a corner, for example...
It doesn't change anything),
I mean, classical TS sometimes even the basis for a normal TS is unsuitable - all relatively true.
Yuri said variance filtering is probably an art like integrating or finding solutions to complex hypotheses.
But also this art is the subtlest subtlety(line) between low PF and high PF.
Yuri said variance filtering is probably as much an art as integrating or finding solutions to complex hypotheses.
But this is also the art of the subtlety (the line) between a low PF and a high one.
No art. I have a dumb algorithm that does everything in a few steps:
- Finding a strategy, i.e. entry conditions - trading signals by indicators and oscillators included in the MT4 package. Custom and other exotics are not used. With the exception of LRMA, because it is very quickly calculated based on two other standard ones. There is no exit, but there is a reversal on the opposite signal. The trading system will always be in the market.
- Getting a list of strategies that give a positive result in a representative sample.
- We go through the list and delete all the MTS that gave less than 100 trades in the representative sample.
- We go through the rest of the list and test all of the MTS for forwards (in my case, the representative sample is the latest data, while out of the sample - the history preceding the representation). No further optimization before the test. If OOS shows negative result, we select next strategy from the list and delete this one.
- A strategy that gives a positive result on OOS, we additionally run it through other financial instruments. No additional optimization before the test. If the result on the majority of additional instruments is negative, we delete the MTS.
As a result of all passing stages, we have one or two MTCs remaining in the list, or most often none at all. If something remains, we slightly optimize the +5/-5 and send it for further testing on demo accounts.
Everything is done by computer, except installing on demo. No art, but a completely mechanical process, like in production: creation, rejection and fine tuning of what was rejected.
IMHO the pattern should be stated in a single sentence understandable to the uninitiated. Until such a level of understanding is reached, and something there is being worked out, adjusted, etc., it is only a matter of time before it is lost. A robust idea can be optimised for the current market, the rest is tweaking.
And what some are trying to automate the search for systems is nonsense and fitting. Only brains can generalize and make new conclusions that are not algorithmized in advance)))) for now...
IMHO the pattern should be stated in a single sentence understandable to the uninitiated. Until such a level of understanding is reached, and something there is being worked out, adjusted, etc., it is only a matter of time before it is lost. A robust idea can be optimised for the current market, the rest is tweaking.
And what some are trying to automate the search for systems is nonsense and fitting. Only the brain can generalize and make new conclusions that are not algorithmized in advance)))) for now...
Delusion is your IMHO generalised by your inflamed brain.
Of course it's easier to flub in a forum than to automate the process.
You can of course look for patterns manually. Nobody forbids it. And you can wash gold manually in a bowl, bit by bit, or you can automate the process with a dredge.
Bullshit is your IMHO, generalised by your inflamed brain.
Of course it's easier to flub in a forum than to automate the process.
Miracle in a sieve, you're the one who's floundering. Any brute force algorithm to find a solution to a story is a shitty fit ;) To analyse intermediate results and on the basis of it set new tasks to collect and analyse statistics is a task for the intellect.
What do you say to a pattern?
Well according to the C.B. market if Williams is overvalued on Thursday there is more than a 50% chance of a black candle on Friday
I've known about it for a long time. >> now I don't know if it works or not. You can draw the rest conclusions...
Miracle in a sieve, you're fluffing. Any brute force algorithm to find a solution to a story is a shitty fit ;) Analyse the intermediate results and based on it set new goals for collecting and analysing statistics is a task for the intellect.
Stay with your IMHO. I honestly don't give a shit. I won't change my mind.
At least while I'm ranting in the forum, the computer is doing most of the work. Slowly looking for strategies, creating ready-made MTCs, checking them for lousiness. I have a lot of free time now. As a last resort, I can manually double-check the results given by the computer. But so far I don't see the need in it - the market is the best inspector, it objectively dots all the i's, i.e. equity.
I used to have to do the same thing manually.
Good afternoon!
Read your thread at my leisure and decided to respond a little:
There are regularities everywhere!!!
You just need to be able to discern them.
In relation to forex, we can say the following:
Let's say on a certain day of the week the closing of the day is higher than its opening in most cases.
On another day it is vice versa.
But this is only the first step!
The second step, we can see how this dependence depends on whether the opening of the current day was higher or lower than the previous one.
After that we can add the number of week in a month, month in a year.
We can also make some more magic.
Please, find enclosed an Expert Advisor with which you can check the first two points and the results of EURUSD analysis for the whole available history.
This EA was made at short notice and there might be some errors.
Any brute force algorithm that tries to find a solution on history is a bullshit fitting ;)
What a pearl!
Where else can you look for solutions and patterns but in history?
The 1st minute bar that closed and everything after it is HISTORY.
Without considering past market behaviour we are stepping into the abyss.
.
>> And there are already many criteria by which one can reliably distinguish
between optimization and fitting has been worked out already.