Machine learning in trading: theory, models, practice and algo-trading - page 3539
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
The topic can be rolled up in laminate, for posterity.
I don't think so.
P.Z.
There's a lot of work to be done here.
Enough for ten years.
If the three of us work together.
I guess.
P.Z.
Usually after my posts,
there's usually a lot of simmering.
I hand over the authority of the topic leader to you and leave in a curtsy ))
Do you want to lure a girl?
And cheat?
P.Z.
Oh, those boys.
Jokers.
I wondered if it is possible to train AMO on indicators and oscillators to restore OHLC prices.
I submitted about 40 indicators and oscillators, trained linear regression....
It restored perfectly.
and this is how it looks like to restore prices from only two signs RSI and SMA
Hasn't anyone realised the interesting properties of the restored prices?
I wondered if it is possible to train AMO on indicators and oscillators to restore OHLC prices.
I submitted about 40 indicators and oscillators, trained linear regression....
It restored perfectly.
and this is what price recovery looks like from only two indicators RSI and SMA.
Indicators are built on the price series, according to some algorithm. Reconstructing a price series from indicators is essentially finding an inverse algorithm. For example, let there is a series of numbers, we make an indicator that, let's say, squares each member of the series and then try to restore the original series by it, i.e. find an algorithm for extracting the root. Naturally, this is not a difficult task. Therefore, imho, it is not surprising that they have recovered perfectly well. But if it were possible to model the future price series, at least for half a period of the indicator, with an error less than that of a naive forecast, it would be very surprising.
.
Doesn't anyone recognise the interesting properties of restored prices?
I wondered if it is possible to train AMO on indicators and oscillators to restore OHLC prices.
I have submitted about 40 indicators and oscillators, trained linear regression....
It restored perfectly.
and this is what price recovery looks like from only two indicators RSI and SMA.
Did no one realise the interesting properties of the restored prices?
At least by eye it is noticeable that the direction is restored with very high accuracy, probably in more than 90% of candles.
Explain.
The first idea is to exploit the lag and sluggishness of indicators.
That is, add a small value up, down and unchanged.
These 3 variants are run through this trained AMK, and if it shows all three variants - a candle down, then as a signal