A poll on what algorithm to run first at lower eval budgets #364

aabbtree77 · 2026-08-01T22:03:32Z

aabbtree77
Aug 1, 2026

It would be great to know which optimization algorithm community prefers for lower function evaluation budgets. Leave your poll answer, do not hesitate to comment or suggest something not listed.

What would you use to optimize an expensive function (e.g. wind turbine process)? Say 100D..1000D evals, D=50.

Default CMA-ES (2001-, most tested algorithm on the planet, frugal in evals with lambda ~log(D))

0%

SLSQP + CMA-ES (Nikolaus Hansen, GECCO 2019)

100%

TREGO (2022, significantly improved EGO/EI which was not easy)

0%

DTS-CMA-ES (2016, could be any surrogate assisted CMA-ES)

0%

RDEx-SOP (winner of CEC-2025)

0%

Something modern with reinforcement learning

0%

Something else

0%

The poll makes no sense

0%

1 vote

Navigation Menu

Uh oh!

A poll on what algorithm to run first at lower eval budgets #364

Uh oh!

{{title}}

Uh oh!

Replies: 0 comments

Select a reply

Uh oh!

Uh oh!

A poll on what algorithm to run first at lower eval budgets #364

Uh oh!

aabbtree77 Aug 1, 2026

Replies: 0 comments

aabbtree77
Aug 1, 2026