Figure 2 | Scientific Reports

Figure 2

From: Combining multi-objective genetic algorithm and neural network dynamically for the complex optimization problems in physics

Figure 2

The distribution of selected individuals which is obtained by using accessibility algorithm. 1000 nondominated individuals are set and 33 individuals are selected. There are two objectives: F1 and F2, and both should be minimized. When the \(\overrightarrow{H}\) is set as (0, 0), it means no preference in F1 and F2, and the distribution of selected individuals is very uniform. When the \(\overrightarrow{H}\) is (20, 0), the individual with lower F1 score prefers to be taken. When the \(\overrightarrow{H}\) is (20, 100), although preference value of F1 is 20, but F2 has higher preference value, and more individuals good at F2 are taken.

Back to article page