Machine learning in trading: theory, models, practice and algo-trading - page 2775
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
What about the fractals, nobody's worked it out? And informative markup.
Especially for this thread, now for half an hour, using only OHLC, roughly sketched an arrow indicator.
This is the first preview without filters and no other tricks only OHLC. This algorithm will work on any TF.
Like it or not, but no R-key and Python will not help if you do not understand the depth and meaning of financial charts. Sorry for the rudeness, of course.
Especially for this thread, now for half an hour, using only OHLC, roughly sketched an arrow indicator.
This is the first preview without filters and no other tricks only OHLC. This algorithm will work on any TF.
Like it or not, but no R-key and Python will not help if you do not understand the depth and meaning of financial charts. Sorry for the rudeness, of course.
It's unscientific, where's the backtest?
You can do it in SQL, the main thing is proof.Differences / increments, velocities / accelerations, spreads / corridors and their averages.
I can't think of anything else to measure.(
Indicators ....
Nothing else is needed in the list.
Everything will come down to a fast enough algorithm that determines the degree of influence of a feature on the target variable. Once again, the algorithm is NOT correlation and NOT the importance of attributes in MO algorithms.
The quality of a trait is the degree of variability in its influence on the target variable.
If you find sufficiently influential attributes (having a high predictive power in my terminology) with less than 10% variation in the numerical estimation of this influence, you will be happy. The prediction error of almost any MO algorithm will be guaranteed to be less than 40%, or you may get lucky and catch up to 20%.
That's the whole plan
It's unscientific. Where's the backtest?
You can do it in SQL, it's all about the proof.Why backtest?
This is a live current chart without peeking into history, as you can do on a tester.
The financial market is alive and slippery like an eel. It requires a special approach.
And you can do scientific work all your life. What's the use? )) I see it, I watch it, I laugh.))
Why the backtest?
It's a live current chart without peeking into history like you can do on a tester.
The financial market is alive and slippery like an eel. It needs a special approach.
And you can do scientific work all your life. What's the use? )) I see, I watch, I laugh.))
and you can't peek into history on a live graph?
and you can't peek at history on a live chart?
I meant peeking into the history one step ahead. It can be done in the tester, which some people find it necessary to take advantage of for personal gain. Not you. I've been watching, you have a good algorithm. I like this approach.
I just made this indicator on my knees. Honestly.
You can set any instrument and TF. We'll test it.
The most difficult thing is data manipulation, you need to learn Pandas or another analogue, then almost any strategy will not be a problem. Or make selections by conditions through SQL queries. Both work quickly.
In my case even loops are not used. It calculates instantly.
Don't think that I'm putting arrows on the graph by hand)))
Especially for this thread, now for half an hour, using only OHLC, roughly sketched an arrow indicator.
This is the first preview without filters and no other tricks only OHLC. This algorithm will work on any TF.
Like it or not, but no R-key and Python will not help if you do not understand the depth and meaning of financial charts. Sorry for the rudeness, of course.
Where does the deal open?
on ovpen where the mark is or on cloze where the mark is or on the next candle?
Where does the deal open?
on ovpen where the mark is or on cloze where the mark is or on the next candle?
I haven't done the scientific work yet.))
Busy soaking such a discovery.)))
P.s.
In principle, it does not matter, it all depends on the TF and the value of stoploss.
Even tried to enter.