Discussing the article: "Brain Storm Optimization algorithm (Part II): Multimodality"

New comment
 

Check out the new article: Brain Storm Optimization algorithm (Part II): Multimodality.

In the second part of the article, we will move on to the practical implementation of the BSO algorithm, conduct tests on test functions and compare the efficiency of BSO with other optimization methods.

In the first part of the article , we delved into the world of optimization with the Brain Storm Optimization (BSO) algorithm revealing the basic principles of this innovative brainstorming-inspired method. Along with studying its logical structure, we also delved into a discussion of clustering methods - K-Means and K-Means++. Brain Storm Optimization (BSO) is an optimization method that incorporates idea generation and evaluation phases in group activities. This algorithm contributed to the field of optimization in conjunction with clustering methods. Clustering allows us to identify groups of similar data elements, which helps BSO find optimal solutions. The use of the mutation method allows the algorithm to bypass obstacles in the solution search space and search for more efficient paths to the optimum.

Now it is time to move on to real action! In the second part, we will dive into the practical implementation of the algorithm, talk about multimodality, test the algorithm and summarize the results.

Author: Andrey Dik

 
If multimodality works, it should show a lot of sine vertices.
 
An example of when a monomodal AO fails.

Forum on trading, automated trading systems and testing trading strategies

Discussion of the article "The role of random number generator quality in the efficiency of optimisation algorithms"

fxsaber, 2024.04.01 19:17

Took a function like this. 
input double X = 0;

double OnTester() { return(MathTan(X)); }

Some obscure result. If you implement iterative poking, I suppose you can find a lot of "rocks".

Tangent is an unsuccessful FF, TS-FF is much easier to poke out.

 
fxsaber #:
If multimodality works, it should show a lot of sine vertices.
I must say that I am not satisfied with the performance of the algorithm as far as multimodality is concerned. In the article I encourage readers to join the research of the algorithm, I think there is a potential for its improvement. Perhaps it is necessary to keep a separate "reference" modal map, so that it could be periodically updated and replenished in the process of optimisation.
 
In my opinion, ESG is much better at modding than ESG, if we go purely visual. I would like to add the possibility for social groups to participate in clustering. Thoughts out loud.
New comment