KI-Glossar
Das vollständige Wörterbuch der Künstlichen Intelligenz
Non-dominated solution set
Collection of solutions where no solution can be considered better than another according to the Pareto dominance criterion. This set represents the approximation of the Pareto front obtained by the optimization algorithm.
Vector fitness function
Extension of the classical fitness function to multi-objective problems, returning a vector of values instead of a single scalar value. This function simultaneously evaluates the performance of a solution across all considered objectives.
Spacing (Spacing metric)
Measure of the uniform distribution of solutions along the approximated Pareto front, calculated from distances between neighboring solutions. A low and regular spacing indicates good diversity and coverage of the Pareto front.
Multicriteria decision making
Process of selecting a final solution from the set of Pareto-optimal solutions based on the specific preferences of the decision maker. This step transforms the optimization problem into a final choice problem.
MOEA/D
Multi-objective decomposition approach that transforms the problem into several single-objective optimization subproblems solved simultaneously. MOEA/D uses neighborhood information to cooperate between subproblems and efficiently explore the Pareto front.