Machine learning in trading: theory, models, practice and algo-trading - page 571
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
So it has little effect. without imports chip, how to understand? ) only if you shuffle the values of each of the predictors in turn, and retrain and watch how the overall error changes
Only a surprise may occur that randomly shuffled data suddenly turn out to be very important :)
Well to mix is to make noise. Noise should compensate itself on average, i.e. go to 0 by weight. I think you can just remove the predictor and get a similar score for it - i.e., important or not.
or this, but still not time efficient
Or so, but still not time efficient
A description of Breyman's Gini forest importance estimation algorithm:
100% per annum is not enough, you need it per month :)
For 100% per annum and stability you will be hired as a trader in any asset management fund on excellent terms, just imagine what kind of capital they have there and what interest you will get. There is enough money for a dozen of Bentleys, but most likely your employers will organize a raid on your apartment, take away your PC and all disks with flash drives, well, to get hold of the technology. And for 300% of the profit the capitalists will kill you without even blinking an eye. So you have to hide everything well, or they will steal the PC and that's it, there will be nothing left of the development.
On the other hand the NS-psychopathy is treated only so - by forced removal of the PC and clicking the switch in the switchboard)))
then as at the very beginning of the thread - calculate in ns and measure the input weights
By the way, for ns, you can also look up how this is defined, exactly...maybe even better
at least for MLP
For 100% per annum and stability you will be hired as a trader in any management fund on excellent terms, imagine what kind of capital they move there and what interest you will have on it. There is enough money for a dozen of Bentleys, but most likely your employers will organize a raid on your apartment, take away your PC and all disks with flash drives, well, to get hold of the technology. And for 300% of profit to the capitalists you will be killed without even blinking an eye. So you have to hide everything well, or they will steal the PC and that's it, there will be nothing left of the development.
On the other hand the NS-psychopathy can be cured only this way - by forcefully taking out the PC and clicking the switch in the switchboard)))
you have a very vivid imagination :)
treated by adjusting the scales in the NS
For 100% per annum and stability you will be hired as a trader in any management fund on excellent terms, imagine what kind of capital they move there and what interest you will have on it. That would be enough for a dozen of Bentleys, but most likely your employers will organize a raid on your apartment, take away your PC and all disks with flash drives, well, to get hold of the technology. And for 300% of profit to the capitalists you will be killed without even blinking an eye. So you have to hide everything well, or they will steal the PC and that's it, there will be nothing left of the development.
On the other hand the only way to cure NS-psychopathy is to forcefully take out the PC and flick the switch in the switchboard)))
To prove the ability to earn 100% per annum, you need at least 2-3 years to show it on your real account.
I did not reach such heights because I do not trade very regularly. I have not reached such heights, because I do not trade very regularly. But to make a minimum of 4-5% per day in manual trading is not a problem. On the machine is less - about 12-16% per month, with daily work.
There is only one problem here. It is only possible with a small deposit. If it increases, the effectiveness of trade falls, and trade itself requires other methods and strategies.
SZZ should be corrected 12-16% on 72 - 96%, because 12-16% / month - this is the profit of the contract amount, ie, without leverage ~ 6. Yes, and we were talking about futures trading on the exchange.
There is only one problem here. It is only possible with a small deposit. With its increase the efficiency of trading falls, and trading itself requires other methods and strategies.
Divide a big deposit into smaller ones. You trade on one account and copy trades on the others.)