Machine learning in trading: theory, models, practice and algo-trading - page 2780
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
Thank you for the systematic presentation of information.
Yes, I have spent a lot of time on my own systematisation of waves. A lot of people don't understand this topic, but it works steadily. Then I moved on to OHLC. There, too, I found a lot of interesting systematic information. The rest is trifles in unification and formation of TS. The MO is interesting in terms of further cognition and revealing the regularities of markets, and more precisely the results of the world economy in the form of charts. There are so many interesting things there, I can't tell you. Doesn't anyone see it? There is no one to seriously discuss it with.))))))
It's one thing to see it, and another thing to write/match the code.
For the hundredth time, by the degree of informational connection
Is mutual information suitable for this?
https://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.mutual_info_classif.html
In classical systems the problem is insufficient flexibility, in MO the problem is excessive flexibility. It is necessary to select data and retrain both. Even the sample size and training frequency are +- the same. Only MO requires many times more power and a "black box". On Onyx back in 2010 they were shoving everything into the grid, since then the capacities have grown by orders of magnitude, but still there.
Why is everyone digging deep when everything is on the surface, on the charts?
Of course, there is no perfect constancy of the exact places, for example, price reversals. There will never be any. But the predictability of price behaviour is not lost because of it. Accuracy can fall, but not predictability. There are interrelated models of markets and there is no escape from them ...
Why does everyone dig deep when everything is on the surface, on the charts?
Of course, there is no perfect constancy of exact places for price reversals. There will never be. But the predictability of price behaviour is not lost because of it. Accuracy can fall, but not predictability. There are interrelated models of markets and there is no escape from them ...
I am generally in favour of manual trading... You can start throwing slippers.
I'm generally in favour of manual trading..... You can start throwing slippers.
I got it. I'm done. I'm done. I'm off to get the slippers.)
Is mutual information suitable for this?
https://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.mutual_info_classif.html
Yeah, aka 21st century correlation.
or http://www.explor edata.net/Yeah, it's a 21st century correlation
Or http://www.exploredata.net/.which is better? this option or the one in scikit-learn?
https://minepy.readthedocs.io/en/latest/python.html
What's better? This one or the one in scikit-learn?
https://minepy.readthedocs.io/en/latest/python.html
both are good, minepy is more advanced, I used it a long time ago, I don't remember the differences.
I don't really support the approach of selecting from a bunch of meaningless features by means of mutual information, rather for quick evaluation of TC norms.
I would even try to put it into an optimiser, as part of a combined optimisation criterion for those who race through genetics.
I don't see any prospects with you. I'm sorry.