This thesis is a study of the notion of "facial attractiveness" in the machine_x000D_ learning with the SVM context. To this end, we have used statistical evaluations_x000D_ made to a group of people who rated front face photographs of women qualifying_x000D_ each on a rating scale according to its appeal. The machine learning technique_x000D_ used is Support Vector Machine (SVM). Then the images were trained in SVM on_x000D_ its respective degree of beauty. By using Chi-square test, we compare the_x000D_ distribution of scores of human evaluators with SVM results show that there is no_x000D_ significant difference between the two groups. The results clearly show that facial_x000D_ beauty is a universal concept, which can be learned by a machine. Analyzing the_x000D_ accuracy of the machine in terms of beauty prediction as a function of the training_x000D_ data size indicates that a mach...