This paper proposes the hybrid evolutionary algorithm (EA) to effectively explore search space by the multiple solutions, each of which changes PSO (Particle Swarm Optimization) and DE (Differential Evolution) temporally and partially. Concretely, the proposed hybrid EA checks whether the solutions can be updated according to the selected PSO or DE as the search strategy, and changes the strategies of the solutions from PSO to DE or vice versa when the p-best is close to the g-best in PSO or when the variance of the solutions becomes small in DE. To investigate the effectiveness of the proposed hybrid EA, this paper compares its performance with that of PSO, DE, and DEPSO-APA as the conventional DE and PSO hybrid EA in the four types of the evaluation functions (i.e., Rosenbrock's Function, Griewank's Function, Rastrigin's Function, and Scaffer's F6 Function). The intensive experiments have revealed the following implications: (1) the proposed hybrid EA can find better solutions than PSO, DE, and DEPSO-APA in Rosenbrock's Function, Rastrigin's Function, and Scaffer's F6 Function and slightly worse in Griewank's Function; and (2) the combination of the change from PSO to DE and from DE to PSO as the core mechanism of the proposed hybrid EA increases not only the possibility of getting out of the local optima but also the local search pressure.

Characterizing an evolutionary algorithm's behavior and performance is a step towards having tools to automatically select and configure the algorithm that better solves the problem at hand. A promising way to characterize algorithms is to use models that capture their dynamics. Dynamic compartmental models are inspired by epidemiology models to study the dynamics of multi- and many-objective evolutionary algorithms. These models have been used as a tool for algorithm analysis, algorithm comparison, and algorithm configuration, assuming that the Pareto optimal set is known. In this work, we relax this assumption by considering the most recent non-dominated set and propose features that allow the use of dynamic compartmental models on large problems. We then introduce a model to estimate the hypervolume from the changes observed on non-dominated solutions in the population. We use several instances of MNK-landscapes with 3, 4, and 5 objectives, and we show that the models can effectively learn algorithm behavior and estimate the search performance of a multi-objective algorithm on those instances. We also show that the models produce good estimates on unseen instances of the same class of problems, and capture the variability of the algorithm when initialized with different populations.