Machine learning in trading: theory, models, practice and algo-trading - page 624
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
At startup the model is trained for 3000 bars (you can see that the green and red lines coincide very well, the variance is minimal). Everything else is forward. And you can see that the model is not so good at predicting already, because of that the variance has increased significantly... but nevertheless I am still surprised that it fits very well on the forward one visually. I don't know if it will work for pair trading, I need to check it.
In the video put the wrong pair, franc vs euro instead of the pound... for the pound vs euro is already worse... but you can try to improve
A couple of charts below:
In the screenshot, the red ones are just increments with a lag of 55 of the current pair,
What does it mean?
The difference between the price of the 0th bar and the price of the 55th bar?
What does it mean?
The difference between the price of the 0th rar and the 55th rar?
log(close[0]/close[55])
You can take the difference instead of the quotient... or you can convert it any way you want. Yuri wrote about the difference between 2 mashcans. You can subtract the mask from the prices.
Then you can reconstruct the original series, as far as I understand
log(close[0]/close[55])
then you can restore the original series, as far as I understand
Why do you take it so far? I thought it was necessary to take with the neighboring...