Neural networks. Questions from the experts. - page 22
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
No, it is an attempt to use non-linear transformations (neural networks) to find patterns in financial timeseries.
No, it is an attempt to find patterns in financial and time series using non-linear transformations (neural networks).
Question ¹2 whether with the help of multilevel neural networks to create a system of verification. For example, take yesterday's graph and compare it with the entire history of this pair. And the results show the following: the segment coincides with the segment of the graph at 1m 10%, at m5 12%, at m15 40%, etc.
and then everything below 50% is filtered out.
can it be done?
Question ¹2 whether with the help of multi-level neural networks to create a system of verification. For example, take yesterday's graph and compare it with the entire history of this pair. And the results show the following: the segment coincides with the segment of the graph at 1m 10%, at m5 12%, at m15 40%, etc.
and then it filters everything below 50%.
can it be done?
We don't need grids for this task. It is easier to calculate the correlation.
I want to use neural networks to impoverish several trading systems into one. I want to use neural networks to impoverish several trading systems into one. for instance, there is a system that earns well in a trend but loses on a flat. then there is the opposite one. then there are those designed for long positions and so on.
For example, if a trend starts, one block is disconnected, another one is added to it. As we don't know the future, delays may occur, and the third block reacts to compensate for delays.
Question ¹2 whether with the help of multi-level neural networks to create a system of verification. For example, take yesterday's graph and compare it with the entire history of this pair. And the results show the following: the segment coincides with the segment of the graph at 1m 10%, at m5 12%, at m15 40%, etc.
and then it filters everything below 50%.
can it be done?
and all of this has to compensate each other. and neural networks have to check the signals of each block, look for patterns. and fully control and learn.
Is a 4-core processor enough for that?
You don't need neural networks for that.
This can be done without NS. In Excel, for example, the Euclidean measure of the distance between a given segment and all - any - other segments is calculated. Then those segments are selected which are the most similar to the given one. The main thing is to remember to scale the data so as not to compare, for example, price fluctuations at 1.4 with fluctuations at 1.2.
You don't need neural networks for that.