An Evolutionary Approach to Automatic Kernel Construction

Howley, Tom; Madden, Michael G.

doi:10.1007/11840930_43

Cited by 18 publications

(23 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another point to note here is that in OSVMs, the kernels that have been used mostly are Linear, Polynomial, Gaussian or Sigmoidal. We suggest it would be fruitful to investigate some more innovative forms of kernel, for example Genetic Kernels [47], that have shown greater potential in standard SVM classification. In the case where abundant unlabeled examples and some positive examples are available, researchers have used many two-step algorithms (as have been discussed in Section 3.3).…”

Section: Discussionmentioning

confidence: 98%

A Survey of Recent Trends in One Class Classification

Khan

Madden

2010

Lecture Notes in Computer Science

Self Cite

278

191

View full text Add to dashboard Cite

Abstract. The One Class Classification (OCC) problem is different from the conventional binary/multi-class classification problem in the sense that in OCC, the negative class is either not present or not properly sampled. The problem of classifying positive (or target) cases in the absence of appropriately-characterized negative cases (or outliers) has gained increasing attention in recent years. Researchers have addressed the task of OCC by using different methodologies in a variety of application domains. In this paper we formulate a taxonomy with three main categories based on the way OCC has been envisaged, implemented and applied by various researchers in different application domains. We also present a survey of current state-of-the-art OCC algorithms, their importance, applications and limitations.

show abstract

Section: Discussionmentioning

confidence: 98%

A Survey of Recent Trends in One Class Classification

Khan

Madden

2010

Lecture Notes in Computer Science

Self Cite

278

191

View full text Add to dashboard Cite

show abstract

“…The macro level is a GA that evolves the kernel shape and parameters for a multiple kernel function. The steady-state genetic algorithm(SSGA) [7][12] is used as underlying mechanism for GA implementation. The micro level algorithm is a RVR algorithm used for computing the fitness of each GA individual from the macro level.…”

Section: B Genetic Evolution Methods Of Complex Multiple Kernel Functionmentioning

confidence: 99%

“…However these researches just selected the multiple kernel(MK) by experience and did not solve how to design a suitable MK automatically. Several recent studies [7][8] [9] have proposed optimization methods which used a genetic algorithm for optimizing the SVM kernels and parameters, but these methods inherit the drawbacks of the nonprobabilistic and Mercer's kernel limits of SVM. This paper uses similar optimization methods to solve the model selection of relevance vector regression.…”

Section: Introductionmentioning

confidence: 99%

Genetic complex multiple kernel for relevance vector regression

Zhang

et al. 2010

2010 2nd International Conference on Advanced Computer Control

View full text Add to dashboard Cite

Relevance vector machine (RVM) is a state-of-theart technique for regression and classification, as a sparse Bayesian extension version of the support vector machine. The selection of a kernel and associated parameter is a critical step of RVM application. The real-world application and recent researches have emphasized the requirement to multiple kernel learning, in order to boost the fitting accuracy by adapting better the characteristics of the data. This paper presents a data-driven evolutionary approach, called Genetic Complex Multiple Kernel Relevance Vector Regression (GCMK RVR), which combines genetic programming(GP) and relevance vector regression to evolve an optimal or nearoptimal complex multiple kernel function. Each GP chromosome is a tree that encodes the mathematical expression of a complex multiple kernel function. Numerical experiments on several benchmark datasets show that the RVR involving this GCMK perform better than not only the widely used simple kernel, Polynomial, Gaussian RBF and Sigmoid kernel, but also the convex linear multiple kernel function.

show abstract

“…While some standard kernels proposed in the literature are straightforwardly used in several applications, tailored kernels produce much better results as each problem has specific characteristics [42,43]. In order to achieve an automated machine learning approach, several works in the literature pose the kernel selection as a search problem in the space of kernels with no human intervention [25,12,28].…”

Section: Introductionmentioning

confidence: 99%

“…Once the space of possible kernels has been defined, the next relevant question is the selection of a strategy to carry out the search. Most of the works in the literature have proposed various heuristic algorithms to solve this search problem, Genetic Programming (GP) being one of the most used methods [25,12,28]. However, there is a lack of knowledge about many aspects related to the specific characteristics of the kernel function optimization problem, and in particular, about how these characteristics relate to the way the GP search for optimal solutions is accomplished.…”

Section: Introductionmentioning

confidence: 99%

In-depth analysis of SVM kernel learning and its components

Roman

Santana

Mendiburu

et al. 2020

Neural Comput & Applic

View full text Add to dashboard Cite

The performance of support vector machines in non-linearly-separable classification problems strongly relies on the kernel function. Towards an automatic machine learning approach for this technique, many research outputs have been produced dealing with the challenge of automatic learning of good-performing kernels for support vector machines. However, these works have been carried out without a thorough analysis of the set of components that influence the behavior of support vector machines and their interaction with the kernel. These components are related in an intricate way and it is di cult to provide a comprehensible analysis of their joint e↵ect. In this paper we try to fill this gap introducing the necessary steps in order to understand these interactions and provide clues for the research community to know where to place the emphasis. First of all, we identify all the factors that a↵ect the final performance of support vector machines in relation to the elicitation of kernels. Next, we analyze the factors independently or in pairs and study the influence each component has on the final classification performance, providing recommendations and insights into the kernel setting for support vector machines.

show abstract

An Evolutionary Approach to Automatic Kernel Construction

Cited by 18 publications

References 8 publications

A Survey of Recent Trends in One Class Classification

A Survey of Recent Trends in One Class Classification

Genetic complex multiple kernel for relevance vector regression

In-depth analysis of SVM kernel learning and its components

Contact Info

Product

Resources

About