Análisis comparativo de los algoritmos Eigenface y Fisherface de reconocimiento facial para la seguridad de los sistemas de información.

The following case study investigated the concepts of the main methods used to date for facial recognition, such as the Eigenface and Fisherface algorithms, the reader was introduced to the mathematical principles for understanding these algorithms as are the concepts of PCA and LDA that are dimensionality reduction methods, these methods are widely used in data analysis in classification algorithms and to be able to give predictions. The PCA and LDA methods are very powerful tools that allow analyzing large data sets, in order to analyze patterns, therefore, studying these methods is very important, and their applications are very wide in Data Science, Machine Learning. In this case study, the different metrics on which they are based are analyzed in order to measure performance, accuracy, sensitivity and precision, in order to reveal which of the two algorithms is the most efficient for system security.

• Autore principale: Riofrio Villamar, Steven Gonzalo (author)
• Natura: bachelorThesis
• Pubblicazione: 2023
• Soggetti: Pca; Lda; Eigenface; Fisherface; Vectores propios; Valores propio; Matriz de confusion; Recall; Especificity; Eigenvectores
• Accesso online: http://dspace.utb.edu.ec/handle/49000/14248