Steganography Genetic Algorithm Hyperparameter Tuning through Response Surface Methodology

Abstract


 
 
 Steganography consists of hidding bits of an information source into a host source. In image processing, a common way of doing the hiding process is to break each byte from the message information and embbed into the message bytes in a way that the differences among the original host and the embedded one is minimized. A genetic algorithm can be used to find the proper combination of bits in order to minimize such differences, but some hyperparameters need to be optimized in order to get an optimized performance. This work investigates the application of Response Surface Methodology in order to find the best hyperparameters of a genetic algorithm applied to image steganography.
 
 

Editorial (Português)

“…O terceiro trabalho, intitulado "Steganography Genetic Algorithm Hyperparameter Tuning through Response Surface Methodology", escrito por da Silva, Carvalho e Martins [3], avalia quais parâmetros de um Algoritmo Genético podem influenciar na otimização de ocultação de mensagens em imagens (esteganografia). Com a análise proposta, os autores ainda sugerem a Metodologia da Superfície de Resposta, como forma de melhor ajuste de parâmetros.…”