Machine learning in trading: theory, models, practice and algo-trading - page 1619
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
Hello ! I've been a trader for about 11 years trading manually on Gunn, but I got into neural networks as well. I solved this problem ! I just want to help you, you have the wrong approach to business. I swear to God this is not a joke, here is the author of this article, where I found the answer to all my questions, he just had an error in the code with the advisor and script, I rewrote everything and I have been trading for almost a year on all currency pairs and timeframes, I try not less than 15 minutes. The program gives you a price forecast, how many pips the price will go and where. https://www.mql5.com/ru/articles/830 ... It all works, I just need to rewrite the code, it has an error there. I do not sell anything and do not advertise anything, I just come and read your letters every day, I want to help in any way I can.
There are some left-wing types in such a respected thread. Do not you find?
Yes, skins are not serious, I propose to measure with indicators, mine is almost ready...
*** you'll measure yourself somewhere else....
I'm definitely redundant in this esteemed thread ))
You are a real fool. Look at my name. My name is Mikhail and my surname is completely different. Don't bullshit....
It's a pseudonym, a clone of Yury Reshetov.
The real name and surname are displayed at the sellers, they are registered under the passport.
What doesMihail Marchukajtes sell? No signals on the real or owls. I am sure there was not a single withdrawal of money from the site to this name. But he constantly advertises Reshetov why, which is like "died", or rather reincarnated into Misha, in every post mentions Yuri apontovyu trinket, releases new versions of his tricks. There's a saying "If something is like a duck, it's a duck", so I say that Misha = Jura. That's Yura's crazy way to promote himself.
And that is why there is no videostream, and even if there is, it will be without his face, it is clear why, because he knows how Yury looks.
It's a pseudonym, a clone of Yury Reshetov.
What does Mihail Marchukajtes sell? No signals on the real or owls. I am sure there was not a single withdrawal of money from the site to this name. But he constantly advertises Reshetov why, which is like "dead", or rather reincarnated into Misha, in every post mentions Yuri apontovyu trinket, releases new versions of his tricks. There's a saying "If something is like a duck, it's a duck", so I say that Misha = Jura. That's Yura's crazy way to promote himself.
And that's why there is no videostream, and even if there is, it will be without a face, it is clear why, because Yury's appearance is known.
You should not insult Reshetov's memory with such a stupid comparison.
By the way, what I don't like about boostings is that the recommended tree depth is 7-10.
That is, if we have 100 predictors, and the division there also starts in the middle of each predictor. Then with a high probability we will have 7 different predictors divided in the middle. Maybe 1 or 2 will divide to a quarter, hardly any smaller.
Or in boosting algorithms, the algorithm doesn't work by half division, but in smaller chunks? Does anyone know?
And who uses what tree depth?
I studied CatBoost, so I'll talk about it.
The depth of the tree is recommended 4-6 splits. This is the depth I try in general.
Predictor splitting is done by three different algorithms to choose from. A so-called grid is created.
The results of splitting are interesting to pull out and see for yourself. And what does AlgLib divide the predictors into equal parts when building the tree for the forest?
Target yours, any... I'm a little torn ....
clusters are only needed for one purpose:
Here we found the HTs on the test new , and accepted them as good...
Now on the new data we need to find this TX to apply the model to it, since the model works well only on HTs, and how do we recognize it on the new data? as an option by the cluster number
What do we feed into the input for clustering - all sample predictors or what?
There is no randomness. The best available partitioning of each predictor is selected. Randomness is in the forest when each tree is fed not all predictors, but for example half of the randomly chosen ones.
It learns once. There is no retraining. For trees/forests seem to have no re-learning at all, apparently from the fact that to re-learn quickly enough.
And why the grid? Trees have nodes and leaves.
Selection will be by linear partitioning, and I'm doing sort of discrete selection by a nonsmooth function, if that makes more sense.
The point is not to do the retraining, but that it is better if the predictors reinforce each other, but do not correlate strongly with each other, and you cannot do that with partitioned sampling.
The predictor partitioning grid, it's not NS.