Machine learning in trading: theory, models, practice and algo-trading - page 915
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
Once again, this cannot be the case in principle. Just repeat the experiment with your data in P on 100+ ELM models and you will not find two identical results. Look for an error.
There isn't. The predictions are different, he means it will be the same if you round them to 0 and 1 (buy and sell). There's a hidden layer, a grain, and an act. function in the script that's searched by genetics. Theoretically it is possible that genetics will find the identical combination several times in a row, then the predictions will be the same, but I have never seen such a thing, the result of genetics is always different.
p.s. Michael, don't embellish. I also said that predictors are good, but all the shamanism should be removed. And increase the training period, which is also constantly talked about here in the topic.
There is no such thing. The predictions are different, he means it will be the same if you round them to 0 and 1 (buy and sell). There is a hidden layer in the script, the grain, and the act. function are enumerated by genetics. Theoretically it is possible that the genetics several times in a row will find an identical combination, then the predictions will be the same, but I have never seen such a thing, the result of genetics is always different.
p.s. Mikhail, don't exaggerate. I also said that the predictors although good, but all sorts of shamanism should be removed. And increase the period of training, which here in the thread is also constantly talking about.
By no means exaggerate and yes I was talking about the signals. The fact that they are on the OOS of all networks is the same. It is clear that the value of the network is different. But the end result in the form of signals is the same... And about the shamanism in the data I do not have it. No peeking into the future, I even take data not from the first bar but from the second, there is no repetition. I'm just a guy, that's all. :-)
About the training period... Let's split into two camps and each respect the other's camp. Let's call one "Longers" the other "Shorters" each has minuses and virtues and they both have a right to exist.
Cons:
The "Longers" CU can afford to sag and sit there for a long time.
With "Shorteners" you have to train day after day.... Or once a week.
Pros:
With "Longers" the TC runs long enough that it frees up the researcher.
In "Shorteners" works almost without errors. There are mistakes, but in terms of profit, they are minimal.
Everyone is free to choose their camp. Who else thinks about the pros and cons of both camps???? Please....
Mishenka, why do we need a camp? Let's better choose a ward. Ward #6))
Which one do you think you are? Just wondering...
I have a friend, we were friends when we were kids. I think everyone has a buddy like that.
He walks around banging his head against everything, 10 concussions... They tell him not to hit it, he does, and then he does it again...
You have to understand that the state of consciousness doesn't improve after each "bump".
Do you have a buddy like that, or do you see yourself as one?
Add Rena and a couple more shots to you, and you'd be fucking nuts about "common sense."
Exactly Max, you are so stubborn that it actually prevents you. You know the saying "All is well in war" and if it works, then I do not ask why, etc. I take it as a given and take it for what it is. So, take this fact that there are people on the market who earn short money, and do not sit in the drawdown for half a year with the hope that it will bail out....
By the way, there will be a separate topic in my video for you and others like you. If you don't like what I say, but if you'll listen to me, it will help you, I'm sure ...
So take it for granted that there are people on the market who make money short-term, not sitting in a slump for half a year, hoping that it will bail out....
It's profitable to sit in the drawdown in order to show a beautiful balance chart, and it will look ugly with shorter periods...
In order to show a beautiful graph of the balance, and with shorts will look ugly...
This will be well visible on the funds chart. Short guys do not sit in a drawdown, if they make a mistake, it is insignificant, if the short guy made a big mistake, it is a signal for retraining, but it is better not to allow this, and retrain him in advance... so he was aware of the current state of the market, which changes with every tick...
What's the message? I thought the millennium was long gone.
You thought you'd capture a little dumbness for future generations?
i'd love to laugh. go ahead.
the stupidity that will make me millions...... :-)