Kernel-based multiple criteria linear programming classifier
نویسندگان
چکیده
This paper proposed a novel classification model which introduced the kernel function into the original Multiple Criteria Linear Programming (MCLP) model. MCLP model is used as a classification method which can only solve linear separable problems in data mining. However, the proposed kernel-based MCLP model can deal with non-linear cases. Meanwhile, unlike some other complicated models, this model is effective and easy to understand. A couple of experimental tests were conducted to evaluate the performance of the proposed kernel-based MCLP model compared with the existing methods: original MCLP and SVM. The results show that kernel-based MCLP model is a competitive method in dealing with nonlinear separable classification problems.
منابع مشابه
Knowledge-Based Nonlinear Kernel Classifiers
Prior knowledge in the form of multiple polyhedral sets, each belonging to one of two categories, is introduced into a reformulation of a nonlinear kernel support vector machine (SVM) classifier. The resulting formulation leads to a linear program that can be solved efficiently. This extends, in a rather unobvious fashion, previous work [3] that incorporated similar prior knowledge into a linea...
متن کاملKernel Regularized Multiple Criteria Linear Programming
Based on our proposed regularized multiple criteria linear programming (RMCLP) for binary classification problems, this paper extends this method to treat with nonlinear case. By applying dual theory, we derived the dual problem of optimization problem constructed in RMCLP, and then proved the solution of RMCLP can be computed by the solution of its dual problem, finally, we constructed Algorit...
متن کاملOptimising Multiple Kernels for SVM by Genetic Programming
Kernel-based methods have shown significant performances in solving supervised classification problems. However, there is no rigorous methodology capable to learn or to evolve the kernel function together with its parameters. In fact, most of the classic kernel-based classifiers use only a single kernel, whereas the real-world applications have emphasized the need to consider a combination of k...
متن کاملMultiple Criteria Decision Making Combined with VRP: A Categorized Bibliographic Study
In this research author reviews references related to the topic of multi criterion (goal programming, multiple objective linear and nonlinear programming, bi-criterion programming, Multi Attribute Decision Making, Compromise Programming, Surrogate Worth Trade-off Method) and various versions of vehicle routing problem (VRP), Multi depot VRP (MDVRP), VRP with time windows (VRPWTW), Stochastic VR...
متن کاملAugmented Kernel Matrix vs Classifier Fusion for Object Recognition
Augmented Kernel Matrix (AKM) has recently been proposed to accommodate for the fact that a single training example may have different importance in different feature spaces, in contrast to Multiple Kernel Learning (MKL) that assigns the same weight to all examples in one feature space. However, the AKM approach is limited to small datasets due to its memory requirements. An alternative way to ...
متن کامل