Machine learning in trading: theory, models, practice and algo-trading - page 504
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
I'm going to try to explain it to you, so that even you understand - you take a price series, break it up, build a model and check it.
What do you do if the model on the test piece of the series shows a "bad" result?
That's right! You change the model, check again, etc.
And what happens in the end?
The same fit.
Here's my suggestion - I take ANY piece of the SB series, break it down and build a model with 70-80% .
What difference does it make - SB or price?
No it's not, I think you're not only a boor, you're as dumb as a cork. You can drive as much as you want, but not on the test row. Not the row your model will be tested on.
You work with half a row, I don't give a shit what you do with it, print it out and stick it in your... You don't see the other half of the row at all, get it? You don't see it, you don't see it, remember that, you don't have it! But it checks the result of your model by me, the market or an independent person, BUT NOT BY YOU!!!
Come again?
Right. You tested it - the model is no good.
Then you change the model.
And so on until the model is good on the test sample.
But in essence, what did you do?
You FITTED the model to the training and test sample.
P.S. B.y., and all this was explained already on 4 years ago...............
NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO !!!!!!!!!!!!!!!!
Fucking stupid...
Checked it once on a test set. O_D_I_N.
Next time on another one.
And then what?
Here's a row of SBs, you can "pull" a model on it (on the entire row).
Let's say you don't know that, and you break it into a test sample and a training sample.
I "pull" the same model on the test sample, which adequately "works" on the whole sample (I did not know this, but the whole sample happened to be such).
That test sample shows the same 70-80% model.
And hooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo.ie??????
NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO !!!!!!!!!!!!!!!!
Fucking stupid...
Checked it once on the test set. O_D_I_N.
Next time on another one.
)))))))))))))) Do you want me to show you that row of SBs?
The Nobel Prize in Physics was awarded yesterday. I remember that you have Einstein as a loon, but look - did you measure normally? You never know...)))
https://indicator.ru/article/2017/10/03/za-chto-dali-nobelevskuyu-premiyu-po-fizike-2017/
yes, again a micro earthquake was mistaken for a wave... but I do not know, I will have to think over what is happening at my leisure, it's all too slippery
it is possible that the space is liquid or jelly but it is not sure
You are hopeless, I wish you were in Valhalla as soon as possible.
))) Let me tell you more (just so you understand) - to get 50% prediction accuracy on a test sample of SB, you have to try really hard.
Too lazy to explain - you wouldn't understand anyway, but just in case, make a note, it's only true for a statistically significant set of samples. Well, law of large numbers, etc...
P.S. God, I'm writing about the law of large numbers to a kid who posts pictures from comic books....
To get 50% prediction accuracy on a test sample of SB, you have to try really hard.
You're right, it's much easier to readjust to noise, especially with grid optimizer on some random indices and 1000 bars dataset.
Get lost unclean please, dibs on me.
Let's have a serious talk. I need your professional advice. Who do you think is going to take who - Odin or Thor?????
You're right, it's much easier to overfit to noise, especially with a grid optimizer, on some random indices and a dataset of 1000 bars.
Well, it's not so easy.
Well.
But this is not about us, not about forex traders, we deal 70% on SB as well, the main thing is to be able to trade this strategy with help of properly placed limit orders, take and stop.