Hi there!
I am trying to apply a PSO technique for black-box function optimization.
As the main instruments for now are written in Python - it will be nice if someone can suggest a library/examples where PSO is used and additional requirements are met:
Function to optimize:
f(p1, p2, ..., pn)
p1, p2, ..., pn - parameters to optimize.
The objective function is unknown, the PSO is planned to apply.
For each parameter to optimize we have a set of boundaries - e.g. for p1:
p1_lb
Yes, there are several Particle Swarm Optimization (PSO) libraries available for use in Python. Here are a few popular ones:
Please refer mealpy library for PSO and other various metaheuristic optimization algorithms for global optimization
https://pypi.org/project/mealpy/
Aws Tamimi , I suppose it can't be used because it can't use initial guess. In my case it's obligatory because the problem is quite complex with the large number of factors.
As far as I see, the main problem is your first prerequisite: variable number of parameters. If you can work around that - for example you could fix the number of parameters to 10 for example, and then in the function f use pi = pi_lb as the instruction to "turn off" the parameter (or something to that effect) - then you can use our Python module Indago: https://pypi.org/project/Indago/
- Badges
- Science topic
- Similar topics
- Indexes