Discussion of article "Self-organizing feature maps (Kohonen maps) - revisiting the subject"
It is quite good and very simple to solve. Just initialise the network by randomly selecting inputs from those you will teach.
It is quite good and very simple to solve. Just initialise the network by randomly selecting inputs from those you will teach.
It is not a problem to calculate, draw maps. There is a lot of work on this topic today.
Where and in what capacity to apply? Clustering? There are more reliable methods.
For myself, I have never found a useful application in trading.
But as an exercise in programming, it is probably useful.
Without minimising the author's efforts.
Good luck
It is not a problem to calculate, draw maps. There is a lot of work on this topic today.
Where and in what capacity to apply? Clustering? There are more reliable methods.
So share the information.
Make your question more specific. It's not clear what information you're talking about.
I decided to test the statistics and this is what I got:
And then, everywhere there is a consistent transition of colours from left to right or from right to left, as it is drawn in the colour palette under the picture. And here with jumping over colours.
There are some shortcomings in the implementation of displaying the results..... But even in this form it is a working variant.
I decided to test the statistics and this is what I got:
And then, everywhere there is a consistent transition of colours from left to right or from right to left, as it is drawn in the colour palette under the picture. And here with jumping over colours.
Save the trained grid and post the grid and data for training. I think when analysing the answer will be found how it is possible. Or, alternatively, will find what the bug is.
In general, we need a reproducible example.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
New article Self-organizing feature maps (Kohonen maps) - revisiting the subject has been published:
This article describes techniques of operating with Kohonen maps. The subject will be of interest to both market researchers with basic level of programing in MQL4 and MQL5 and experienced programmers that face difficulties with connecting Kohonen maps to their projects.
Training without a supervisor is exercised in the SOM. For training purposes competition mechanisms are applied. When sending a pattern network to input, the neuron with a vector that is the least different from the input pattern wins. The following ratio applies for the winner neuron:
where:
Euclidean space is most frequently used as distance.
Author: Nikolay Demko