Hyperparameter Optimization on Ensemble Regression Tree for Lip Coloring Simulation

Technology helps us in many activities and keeps growing, so it makes more efficient, time-saving, using fewer resources also information entertainment are accessible. Machine Learning technology is the fastest-growing field computer science that used areas such as marketing, healthcare, manufacturing, security, transportation. One of machine learning methods Ensemble Regression Tree (ERT) which has succeeded detecting facial features on eyebrows, eyes, nose, lips. However, utilization ERT method not been found to detect specific lips only for gaining optimization. Then this research will be conducted extract feature annotation dataset from iBUG 300W with 68 20 lip area points. The results extraction reduced error rate, saving, still detected coloring simulation was successfully carried out configuration hyperparameter values, tree = 4, regularization 0.25, cascade 8, pool 500, oversampling 40 translation jitter 0. From observations discovered optimization hard disk resource savings 69.36%, RAM 30.8%, CPU 3.8%; reduce rate by 0.058%; increase inference speed 39%.