Linear Genetic Programming for Design Graph Neural Networks for Node Classification

In recent years, significant efforts have been made to address graph node classification tasks by applying graph neural networks and methods based on label propagation. Despite the progress achieved by these approaches, their success often hinges on complex architectures and algorithms, sometimes leading to the oversight of crucial technical details. One crucial aspect of the innovative approach in designing artificial neural networks is suggesting a novel neural architecture. Currently used architectures have primarily been developed manually by human experts, which is time-consuming and error-prone. That is why adopting more sophisticated semiautomatic methods, such as Neural Architecture Search, has become commonplace. This paper introduces and assesses a Linear Genetic Programming approach for designing graph neural networks in the context of Node Classification. Our approach aims to systematically define the GCN parameter space, drawing inspiration from recent research on design principles. By doing so, our method seeks to balance achieving satisfactory performance and optimizing memory and computation resources, thus offering a more efficient alternative to conventional approaches from the neural architecture search area.