<![CDATA[This research aim to scoring the facial look by implementing geometrical and texture features using multiple linear regression on scikit-learn library. Geometrical features calculate the range of facial landmark features, while texture features were calculated using a local binary pattern on scikit-image library. The method used for creating the model was multiple linear regression. This method models the relationship between independent variables (features) and dependent variables (score from humans), so the model can be used to predict new data. The score predicted by the machine learning model was compared to the average score from 13 respondents that were collected by a questionnaire. The result of scoring from the machine learning model and questionnaire were evaluated by mean squared error (MSE). The result showed the model can predict the score of facial look close to scoring from humans with MSE value 0,43. This model is built using a subjective opinion label from 13 respondents. Thus, the conclusion drawn from this research only give the big picture of the dataset used, and not a form of reality validation..]]>
Copyrights © 2022
