Machine learning in trading: theory, models, practice and algo-trading - page 3486
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
So, I have trained the random and I hasten to inform you that on the samples train, test, exam the balance was zero, i.e. no regularity was found on all samples, as I understand, so the activation threshold was not exceeded.
Reduced the learning rate, and still managed to get activating patterns, below 3 samples train, test, exam.
The results are not outstanding relative to the non-random and transformed sample.
went back many pages to where the argument was,
I don't understand how AI can work in the market, because where AI succeeds is in a narrow circle, physiological as a rule, where there are reference points and boundaries,
pixels in/n/l/p, if you look at appearance for example, but what can you do with a market that has no physiological boundaries?
only in theory such developments
Yes, it can work, and not even AI, but ordinary neural networks, forests-trees, etc. Bayesian. The market is physiology, i.e. psychology, mass. In the end, it all comes down to trivial statistics. Don't just demand impossible things from neural networks and so on.
for everything there must be a reference point and boundaries, where AI is working now, there is one, and there are boundaries of all areas for it.
Is this possible in the market?
The values in the rows change. If earlier to lines 0-100 corresponded predictor values, for example from 2,55 to 2,88, which are indicated by quantum 0 (values are saved in the table of values by quantum), and 900-1000 were from 3,44 to 3,77. So by replacing the quantum number in these lines you change their values. Which is equivalent to changing the values in the rows of this predictor. You don't move the teacher's values in the same rows. I.e. you teach the wrong thing: 0 line of the teacher you give the value from the 900 line of the predictor.
I'm already tired of explaining. Do as you wish.
Okay, I agree that the predictor value changes, but the target line does not change.
Thanks for trying so hard to help me!
anatomy (in which AI has excelled) does not jump up to the bottom unpredictably, going beyond the limits, everything is strict there, the rest is cut off and given to error
error is not suitable for the market, there are a number of other things that say about the MO that it only works with tangible objects now.
the market is balancing, if the MO helps, it's a very low percentage, low but profitable maybe, not what someone expects from it here maybe.
To fit the unfit.
It's a famous fool's test, where you train on random data and fit it to a test. And then compare it to training on the original data.
If training on randoms fits on average as good or better, you're in trouble. That's p-hacking.
Another disturbing thing is how to understand now what is random and what is regularity.....
Apparently, we need to significantly raise the requirement for the model and TC estimation.
I wonder how to calculate the probability of profitable TS from chaos without intentional regularities and from chaos with some human-made regularities.
Answered your questions. #416.
I understand a little bit about it, as I myself have been puzzling over a similar task for a long time
Apparently there was so much spam that day that all the answers and thoughts were lost....
The other thing that is disturbing is how to figure out now what's random and what's a pattern....
The other thing that is disturbing is how to figure out now what's random and what's a pattern....
Apparently, it is necessary to significantly raise the requirement for the model and evaluation of the TS.
I wonder how to calculate the probability of profitable TS from the chaos without intentional regularities and from the chaos with some human-made regularities.
I didn't really look at the graph. I'm looking at it now.