AUTHOR=Amaya-Tejera Nazhir , Gamarra Margarita , Vélez Jorge I. , Zurek Eduardo 

TITLE=A distance-based kernel for classification via Support Vector Machines

JOURNAL=Frontiers in Artificial Intelligence

VOLUME=Volume 7 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2024.1287875

DOI=10.3389/frai.2024.1287875

ISSN=2624-8212

ABSTRACT=Here we propose an innovative classification method and distance-based kernel for Support Vector Machines (SVMs). In contrast to other classification procedures where the data set is partitioned for training and testing once, we randomly select subsets of the data to train the model a finite number of times and subsequently identify the representative data subset, which leads to better inferences about the population. Computational experiments with publicly available data of different sizes, showed that our proposal improved the accuracy. In addition, the proposed kernel is based on a similarity matrix for binary-type features, which can be applied to binary and multiclass classification problems. We found that our distance-based kernel overcome other well-known kernels in the literature and to other authors using the same datasets.