Abstract
Multi-objective evolutionary optimization has found increasing applications in the real world, many of which are expensive. This chapter starts with introducing three main categories of evolutionary algorithms for multi-objective optimization, namely decomposition based, Pareto dominance based and performance indicator based. This is followed by a description of three representative surrogate-assisted evolutionary multi-objective optimization algorithms, all of which use the acquisition functions in Bayesian optimization, also known as the infill criteria, for model management. The first two algorithms are based on the decomposition based approach, while the third one adopts the non-dominated sorting method as the baseline search algorithm. Different to the first two algorithms, the third one focuses on reducing the computational complexity of Bayesian optimization by replacing the Gaussian process with a heterogeneous ensemble, making it applicable to high-dimensional expensive problems.