Machine learning in trading: theory, models, practice and algo-trading - page 1733
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
Everything works. We need to come up with floating time windows. Fixed numbers seem to be limited.
If there is a "spilled" regularity by minutes, I don't think it's worth clustering, just leave it as it is and test it.
Well, if the pattern is "spilled" unevenly/uniformly over the minutes, then I think it is not worth clustering, for now leave it as it is and test
usually grouped by several minutes in a row.
usually grouped for several minutes at a time
Do it your way, but do OOS.
want to see
This is a question for experts in MO: it is known that you can teach the NS to recognize an item from an image, but the item must be in normal, assembled condition, but is it possible to teach the NS to recognize broken items, such as a car after an accident, a house in the process of demolition, or furniture after a tornado? A human can do it in one go.
What difference does it make whether the house is broken or not, the network learns what it is taught
What difference does it make whether the house is broken or not, the network learns what it is taught
Exhaustive.))
A house is always broken in different ways. There is a big difference between a whole house and a broken house. If a whole house has a few christomatic images, a broken house can look like anything. And yet, one recognizes it easily.
Humans can easily deal with entropy in an image, but NS?
I showed somewhere (I forgot where, because I haven't been trading for more than a month) that the probability distribution of market increments is the product of the CB Gaussian and exponential (or in general, the Erlang distribution) distributions.
The Erlang distribution is responsible for the time intervals between the tick quotes and the generator of such numbers looks like this:
Here Lambda is the intensity of the flow of events (quotes).
If Lambda=const, the process is stationary, but the flux intensity is different at different moments of the market, i.e. Lambda=f(t) that determines the non-stationarity of the process in general.
So, in order to distinguish the stationary process, it is necessary to consider separate sections of BP with the same flux density as a whole.
Therefore, the attempts to divide the BP into hours within a day, and then these hours "glue" together - clearly have the right to life.
Exhaustive.))
A house is always broken in different ways. Between the whole house ...
Well, yes, and the cats in the picture are different, but the network recognizes them and distinguishes them from the dogs somehow ...
Read at least something about the principles of pattern recognition, about convolutional networks, how it works, etc. Your questions are very immature, when you read them you'll understand their stupidity.