Principled Evolutionary Algorithm search operator design and the kernel trick

Lane, Fergal; Azad, R. Muhammad Atif; Ryan, Conor

doi:10.1109/ssci.2016.7850204

Cited by 3 publications

(4 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A second motivation has been our interest in the use of kernel methods [39] in the principled design of Evolutionary Algorithms (EA) and EA search operators [26,27]. At the core of every kernel model is a kernel function that is chosen to match the inherent statistical characteristics of the problem domain at hand.…”

Section: Kernel Methodsmentioning

confidence: 99%

DICE: A New Family of Bivariate Estimation of Distribution Algorithms Based on Dichotomised Multivariate Gaussian Distributions

Lane

Azad

Ryan

2017

Applications of Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

A new family of Estimation of Distribution Algorithms (EDAs) for discrete search spaces is presented. The proposed algorithms, which we label DICE (D iscrete Correlated Estimation of distribution algorithms) are based, like previous bivariate EDAs such as MIMIC and BMDA, on bivariate marginal distribution models. However, bivariate models previously used in similar discrete EDAs were only able to exploit an O(d) subset of all the O(d 2) bivariate variable dependencies between d variables. We introduce, and utilize in DICE, a model based on dichotomised multivariate Gaussian distributions. These models are able to capture and make use of all O(d 2) bivariate variable interactions in binary and multary search spaces. This paper tests the performances of these new EDA models and algorithms on a suite of challenging combinatorial optimization problems, and compares their performances to previously used discrete-space bivariate EDA models. EDAs utilizing these new dichotomised Gaussian (DG) models exhibit signicantly superior optimization performances, with the performance gap becoming more marked with increasing dimensionality.

show abstract

Section: Kernel Methodsmentioning

confidence: 99%

DICE: A New Family of Bivariate Estimation of Distribution Algorithms Based on Dichotomised Multivariate Gaussian Distributions

Lane

Azad

Ryan

2017

Applications of Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

show abstract

“…Kernel trick is a function that takes as inputs vectors in the original space and returns the dot product of the vectors in the feature space is called a kernel function also referred to as kernel trick. It finally returns a similarity score between any two points, which can be considered as a metric of closeness [46,47].…”

Section: Second Featurementioning

confidence: 99%

SWFT: Subbands wavelet for local features transform descriptor for corneal diseases diagnosis

Al-Salihi¹,

Aydin²,

Ghaeb³

2021

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

A human cornea is the front see-through shield of the eye; that refracts light onto the retina to induce the vision. Therefore, any defect in the cornea should lead to vision disturbance. This deficiency is estimated by sets of topographical images measured assessed by an ophthalmologist. Consequently, an important priority is the early and accurate diagnosis of diseases that may affect corneal integrity through the use of machine learning algorithms. Images produced by a Pentacam device can be subjected to rotation or some distortion during acquisition, therefore, accurate diagnosis requires the use of local features in the image. Accordingly, a new algorithm called subbands wavelet for local features transform (SWFT) developed which mainly based on the algorithm of a scale-invariant feature transform (SIFT). This algorithm uses wavelets as a multiresolution analysis to produce images with different scales instead of using the Difference of Gaussians (DoG) as in the SIFT algorithm. The Experimental results on the corneal topography dataset, indicate that the proposed SWFT outperforms the baseline SIFT algorithm.

show abstract

“…One promising exception is the recent study by Lane et al [25]. They propose the use of kernels in the context of evolutionary search operators.…”

Section: Integration I: Search Operators and Kernelsmentioning

confidence: 99%

“…They propose the use of kernels in the context of evolutionary search operators. As the very same kernels may then be employed in the optimization algorithm (here: an evolutionary algorithm) as well as the model (e.g., Kriging or SVMs) Lane et al [25] state that this might lead to a more seamless integration between EAs and kernel-based surrogate models being used to augment them.…”

Section: Integration I: Search Operators and Kernelsmentioning

confidence: 99%

Towards Better Integration of Surrogate Models and Optimizers

Chugh

Rahat

Volz

et al. 2019

High-Performance Simulation-Based Optimization

View full text Add to dashboard Cite

Surrogate-Assisted Evolutionary Algorithms (SAEAs) have been proven to be very effective in solving (synthetic and real-world) computationally expensive optimization problems with a limited number of function evaluations. The two main components of SAEAs are: the surrogate model and the evolutionary optimizer, both of which use parameters to control their respective behavior. These parameters are likely to interact closely, and hence the exploitation of any such relationships may lead to the design of an enhanced SAEA. In this chapter, as a first step, we focus on Kriging and the Efficient Global Optimization (EGO) framework. We discuss potentially profitable ways of a better integration of model and optimizer. Furthermore, we investigate in depth how different parameters of the model and the optimizer impact optimization results. In particular, we determine whether there are any interactions between these parameters, and how the problem characteristics impact optimization results. In the experimental study, we use the popular Black-Box Optimization Benchmarking (BBOB) testbed. Interestingly, the analysis finds no evidence for significant interactions between model and optimizer parameters, but independently their performance has a significant interaction with the objective function. Based on our results, we make recommendations on how best to configure EGO.

show abstract

Principled Evolutionary Algorithm search operator design and the kernel trick

Cited by 3 publications

References 21 publications

DICE: A New Family of Bivariate Estimation of Distribution Algorithms Based on Dichotomised Multivariate Gaussian Distributions

DICE: A New Family of Bivariate Estimation of Distribution Algorithms Based on Dichotomised Multivariate Gaussian Distributions

SWFT: Subbands wavelet for local features transform descriptor for corneal diseases diagnosis

Towards Better Integration of Surrogate Models and Optimizers

Contact Info

Product

Resources

About