Various Degrees of Steadiness in NSGA-II and Their Influence on the Quality of Results

Buzdalov, Maxim; Parfenov, Vladimir

doi:10.1145/2739482.2764887

Cited by 5 publications

(4 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the EMO literature, there exists many algorithms based on the steadystate evolution model (e.g., [9]- [16]). In some recent studies (e.g., [8], [17], and [18]), the steady-state EMO algorithm has shown better performance, in terms of convergence and diversity, than its generational counterparts on some problems. To our best knowledge, most, if not all, studies on the NDS are discussed in the context of a generational evolution model, whereas few have considered the situation for a steady-state evolution model yet.…”

mentioning

confidence: 99%

Efficient Nondomination Level Update Method for Steady-State Evolutionary Multiobjective Optimization

Deb

Zhang

et al. 2017

IEEE Trans. Cybern.

View full text Add to dashboard Cite

Abstract-Nondominated sorting (NDS), which divides a population into several nondomination levels (NDLs), is a basic step in many evolutionary multiobjective optimization (EMO) algorithms. It has been widely studied in a generational evolution model, where the environmental selection is performed after generating a whole population of offspring. However, in a steady-state evolution model, where a population is updated right after the generation of a new candidate, the NDS can be extremely time consuming. This is especially severe when the number of objectives and population size become large. In this paper, we propose an efficient NDL update method to reduce the cost for maintaining the NDL structure in steady-state EMO. Instead of performing the NDS from scratch, our method only updates the NDLs of a limited number of solutions by extracting the knowledge from the current NDL structure. Notice that our NDL update method is performed twice at each iteration. One is after the reproduction, the other is after the environmental selection. Extensive experiments fully demonstrate that, comparing to the other five state-of-the-art NDS methods, our proposed method avoids a significant amount of unnecessary comparisons, not only in the synthetic data sets, but also in some real optimization scenarios. Last but not least, we find that our proposed method is also useful for the generational evolution model. Index Terms-Computational complexity, nondominated sorting (NDS), nondomination level (NDL), Pareto dominance, steady-state evolutionary multiobjective optimization (EMO).

show abstract

mentioning

confidence: 99%

Efficient Nondomination Level Update Method for Steady-State Evolutionary Multiobjective Optimization

Deb

Zhang

et al. 2017

IEEE Trans. Cybern.

View full text Add to dashboard Cite

show abstract

“…In particular, the multi-objective algorithms (NS-LC, SS-LC, NSS-LC, NS-SS-LC) use an improved version of NSGA-II for steady-state implementations [39]. There are three main reasons behind this choice: first, for fair comparisons with the baselines used in this work (NS, SS, NSS) which use a steady state implementation for the maze navigation testbed [2], [7], [20]; second, due to evidence in [46] that the generational counterpart of novelty search does not perform equally well in a maze navigation scenario, likely due to a "less informative gradient" for novelty search given by a generational reproduction mechanism [26]; third, due to arguments in recent studies [47]- [49] that a steady-state multi-objective implementation can be beneficial in terms of convergence and diversity for particular problems. Based on these studies we assume that a steady state implementation is more suitable for the maze navigation testbed.…”

Section: Discussionmentioning

confidence: 99%

Quality Diversity Through Surprise

Gravina

Liapis

Yannakakis

2019

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

Quality diversity is a recent family of evolutionary search algorithms which focus on finding several wellperforming (quality) yet different (diversity) solutions with the aim to maintain an appropriate balance between divergence and convergence during search. While quality diversity has already delivered promising results in complex problems, the capacity of divergent search variants for quality diversity remains largely unexplored. Inspired by the notion of surprise as an effective driver of divergent search and its orthogonal nature to novelty this paper investigates the impact of the former to quality diversity performance. For that purpose we introduce three new quality diversity algorithms which employ surprise as a diversity measure, either on its own or combined with novelty, and compare their performance against novelty search with local competition, the state of the art quality diversity algorithm. The algorithms are tested in a robot navigation task across 60 highly deceptive mazes. Our findings suggest that allowing surprise and novelty to operate synergistically for divergence and in combination with local competition leads to quality diversity algorithms of significantly higher efficiency, speed and robustness.

show abstract

“…In this paper we have used the tree structure for the set of fronts and the solutions inside the fronts are considered in linear manner. It would be interesting to see whether the tree structure in the fronts can improve the number of dominance comparison as done in [22], [30].…”

Section: S O R T I N Gmentioning

confidence: 99%