Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man

Table 5

Win rates for testing results across all 10 runs.

Proposed algorithm	PAESNet_F	PAESNet_V	PAESNet_M
PAESNet_F	—	30%	20%
PAESNet_V	70%	—	40%
PAESNet_M	80%	60%	—

(𝐴 versus 𝐵) + (𝐵 versus 𝐴) = 100%.