Re-sampled inheritance search: high performance despite the simplicity

Caraffini, Fabio; Neri, Ferrante; Passow, Benjamin N.; Iacca, Giovanni

doi:10.1007/s00500-013-1106-7

Cited by 32 publications

(30 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To produce our numerical results, each optimiser was run 30 times with the parameter settings suggested in the original papers, and a computational budget B = 5000D (D = 10, 50, 100 dimensions). As for RI-(1+1)-CMA-ES, the same parameters setting proposed in [6] was used for the "restart" procedure, and the one suggested in [3] for the (1+1)-CMA-ES component. The statistical analysis suggested in [21] was adopted to further validate our conclusions.…”

Section: Methodsmentioning

confidence: 99%

“…To overcome this problem, and obtain an algorithm that may preserve the performance of the original CMA-ES while maintaining a low level of algorithmic and computational complexity, we introduce here a version of (1+1)-CMA-ES that includes the so-called Re-sampling Inheritance Search (RIS), which was proven to be a simple yet efficient framework, in particular on large scale and separable problems [6,7]. This idea follows the Ockham's Razor advocated in [8], which states that simplicity should always be considered a crucial aspect of algorithmic design.…”

Section: Introductionmentioning

confidence: 99%

“…However, it is important to notice that we do not propose the same restart process, but rather an iterated local search similar to RIS. More specifically, not only we reset the covariance matrix, but we also sample a random new solution, and subsequently we feed it to an exponential crossover (see [6] for details), together with the current best solution ("elite") found by the algorithm. The resulting solution then shares some variables with the current best solution, and is used after the restart as new start point for (1+1)-CMA-ES.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Improving (1+1) covariance matrix adaptation evolution strategy: A simple yet efficient approach

Caraffini

Iacca

Yaman³

2019

AIP Conference Proceedings

Self Cite

View full text Add to dashboard Cite

In recent years, part of the meta-heuristic optimisation research community has called for a simplification of the algorithmic design: indeed, while most of the state-of-the-art algorithms are characterised by a high level of complexity, complex algorithms are hard to understand and therefore tune for specific real-world applications. Here, we follow this reductionist approach by combining straightforwardly two methods recently proposed in the literature, namely the Re-sampling Inheritance Search (RIS) and the (1+1) Covariance Matrix Adaptation Evolution Strategy (CMA-ES). We propose an RI-(1+1)-CMA-ES algorithm that on the one hand improves upon the original (1+1)-CMA-ES, on the other it keeps the original spirit of simplicity at the basis of RIS. We show with an extensive experimental campaign that the proposed algorithm efficiently solves a large number of benchmark functions and is competitive with several modern optimisation algorithms much more complex in terms of algorithmic design.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improving (1+1) covariance matrix adaptation evolution strategy: A simple yet efficient approach

Caraffini

Iacca

Yaman³

2019

AIP Conference Proceedings

Self Cite

View full text Add to dashboard Cite

show abstract

“…Evolutionary Algorithms (EAs) are widely used to solve the multi-objective optimisation problem in transportation ( [6], [7], [8], [9], [10], [11]). However, in a traffic signal plan optimization process, traffic simulation needs to be called every time an individual is evaluated.…”

Section: Introductionmentioning

confidence: 99%

Improving anytime behavior for traffic signal control optimization based on NSGA-II and local search

Nguyen

Passow

Yang

2016

2016 International Joint Conference on Neural Networks (IJCNN)

Self Cite

View full text Add to dashboard Cite

Abstract-Multi-Objective Evolutionary Algorithms (MOEAs) and transport simulators have been widely utilized to optimise traffic signal timings with multiple objectives. However, traffic simulations require much processing time and need to be called repeatedly in iterations of MOEAs. As a result, traffic signal timing optimisation process is time-consuming. Anytime behaviour of an algorithm indicates its ability to return as good solutions as possible at any time during its implementation. Therefore, anytime behavior is desirable in traffic signal timing optimisation algorithms. In this study, we propose an optimisation strategy (NSGA-II-LS) to improve anytime behaviour based on NSGA-II and local search. To evaluate the validity of the proposed algorithm, the NSGA-II-LS, NSGA-II and MODEA are used to optimize signal durations of an intersection in Andrea Costa scenario. Results of the experiment show that the optimization method proposed in this study has good anytime behaviour in the traffic signal timings optimization problem.

show abstract

“…In this paper, the initial local search is performed by the so-called Short Distance Exploration or simply S algorithm, see [30], [31], and [32]. The S algorithm is a simple greedy local search that performs moves along the axes and halves its radius when it is unable to detect a better solution.…”

Section: B Super-fit Dementioning

confidence: 99%

Robust impaired speech segmentation using neural network mixture model

Iliya

Menzies²,

Neri

et al. 2014

2014 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT)

Self Cite

View full text Add to dashboard Cite

Abstract-This paper presents a signal processing technique for segmenting short speech utterances into unvoiced and voiced sections and identifying points where the spectrum becomes steady. The segmentation process is part of a system for deriving musculoskeletal articulation data from disordered utterances, in order to provide training feedback. The functioning of the signal processing technique has been optimized by selecting the parameters of the model. The optimization has been carried out by testing and comparing multiple Differential Evolution implementations, including a standard one, a memetic one, and a controlled randomized one. Numerical results have also been compared with a famous and efficient swarm intelligence algorithm. For the given problem, Differential Evolution schemes appear to display a very good performance as they can quickly reach a high quality solution. The binomial crossover appears, for the given problem, beneficial with respect to the exponential one. The controlled randomization appears to be the best choice in this case. The overall optimized system proved to segment well the speech utterances and efficiently detect its uninteresting parts.

show abstract

Re-sampled inheritance search: high performance despite the simplicity

Cited by 32 publications

References 58 publications

Improving (1+1) covariance matrix adaptation evolution strategy: A simple yet efficient approach

Improving (1+1) covariance matrix adaptation evolution strategy: A simple yet efficient approach

Improving anytime behavior for traffic signal control optimization based on NSGA-II and local search

Robust impaired speech segmentation using neural network mixture model

Contact Info

Product

Resources

About