Machine learning in trading: theory, models, practice and algo-trading - page 711
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 Grail is not here, he's sitting next to me, watching this correspondence with bulging eyes.
I'm also slowly getting to all sorts of probability densities and stuff, but a little bit from the other side, from moushilling :) I'm almost there, the only thing left is to understand coolerning more or less in its different manifestations
entropy is used there and different distributions can be set for the agents
I'm also gradually approaching all sorts of probability densities and stuff, but a little bit from the other side, from moushinlerning :) I'm almost there, the only thing left is to understand coolerning more or less in its various manifestations
That's right.
Just deal with entries - they should not be random, but strictly based on trading intensity. In the long run, in the long run.
If there will be some difficulties with finances just for the sake of living - give me a vote, and I`ll just drop the Grail on the probabilities, in order to have strength to continue the epic.
Show must go on!
Open your eyes)))
opened))))
good luck!
)))
That's right.
Just deal with the entries - they should not be random, but strictly tied to the intensity of the trades. For time, in the long run.
If there will be some difficulties with finances just for the sake of living - give me a vote, and I'll just drop the Grail on the probabilities, in order to have strength to continue the epic.
Show must go on!
the main thing would be interest :) while there are interesting unexplored things, you can poke around
I see no sense in supervised and unsupervised lerning, mere approximation and feature search for narrow stationary problems
I'm all for boosting too, but I disagree a bit with his fight against overfit.
For him, defeating overfit = specific tree partitioning criterion + minimum number of trees.
For me: gbm + k-fold. I stopped on gbm now, but tried other packages in R - xgboost, catboost - they have built-in overfit protections, but I didn't get as much joy as with gbm.
Is it?
Anyway, here's the bottom line, the top line I think everyone will figure it out on their own
In contrast to the neural networks, I have shown a higher indent, here from the night to now (of course real):
The curve is of course interesting and can not be ignored, but the yield is too low. IMHO
Is it?
Yes. 1 в 1.
there's the most interesting part, including for DC - the forecast
can not
Teacher, once again you've been figured out by the normal guys)))