Skip to main content
Version: Next

moo_base Overview

This page introduces the moo_base project.
This project optimizes the multi-objective benchmark function ZDT1[9].

optimization.yaml (moo_base project)​

The contents of optimization.yaml are shown below.
The evaluator is the zdt1 implementation of the ZDT1 function, with 30 design variables.
The optimizer is the multi-objective optimization algorithm nsga2 with a fixed random seed. The design variables are bounded by [0, 1].

evaluator:
  name: zdt1
optimizer:
  name: nsga2
  kwargs:
    seed: 42
    bounds: [0.0, 1.0]

This example performs 30 optimization iterations.

num_iteration: 30

(Output-related settings are omitted.)

Optimization Example​

The GUI after completing 30 optimization iterations is shown below. A uniform Pareto front can be observed. According to the literature, this Pareto front is very close to the true Pareto front.
moo_base GUI