Discussion of article "Population optimization algorithms: Harmony Search (HS)"
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
New article Population optimization algorithms: Harmony Search (HS) has been published:
In the current article, I will study and test the most powerful optimization algorithm - harmonic search (HS) inspired by the process of finding the perfect sound harmony. So what algorithm is now the leader in our rating?
The work of HS logic is similar to the work of a musician in creating perfect harmony. The musician tries to change the various tones until the perfect harmony is found. After that, the collection of found harmonies is stored in memory. In an optimization problem, harmonies undergo various changes; if the results of the variation are favorable, then the memory is renewed by adding harmony to the memory and removing the undesirable elements... All this may sound pretty confusing. So what is harmony? What are tones? Let's try to understand the algorithm using our own terms.
What is a piece of music? Of course, I'm not a musician (which is a pity), but a programmer. But for the sake of the algorithm detection, it will be sufficient to apply the concept of "note". A piece of music is composed of notes (chords). Figure 1 schematically shows the "mechanism" for building a piece of music. The selection of notes corresponds to a piece of music, which is easily determined even without an ear for music or musical education. Those willing to guess may leave a comment below.
Optimizing the HS algorithm consists in moving the green bars with notes across the blue bar of the piece itself. The range of the green bar is an octave, which is made up of individual notes. The product (blue bar) corresponds to one of the optimization solutions. The notes on the green bar are the corresponding optimization parameters of the problem. The musician's memory stores several versions of the piece (several variants of blue bars), this is the algorithm population.
Figure 1. Selection of notes in a piece of music (search for harmony). The blue bar is a piece. The green bars are set of notes
The example in Figure 1 corresponds to the solution of a discrete problem, where there are eight steps in the parameter. The example is provided for ease of understanding the operation of the algorithm. However, in an arbitrary task, there can be any step of the optimized parameters and there are also intermediate notes - semitones. The correct parameters for solving the problem correspond to the correct notes in the piece.
Author: Andrey Dik