학술논문

Enhancement of Precision in Facial Age Identification using Ensemble Support Vector Machine Algorithm in Comparison with Lasso Regression Algorithm
Document Type
Conference
Source
2024 4th International Conference on Data Engineering and Communication Systems (ICDECS) Data Engineering and Communication Systems (ICDECS), 2024 4th International Conference on. :1-5 Mar, 2024
Subject
Communication, Networking and Broadcast Technologies
Computing and Processing
Robotics and Control Systems
Support vector machines
Image recognition
Calculators
Communication systems
Face recognition
Data engineering
Vectors
Facial image
Facial Recognition
Innovative Ensemble Support Vector Machine
Lasso Regression
Feature Extraction
facial growth
Age
Language
Abstract
The article compares the Lasso Regression (LR) technique with the Innovative Ensemble Support Vector Machine (ESVM) algorithm to recognise and determine facial age. The image collection is processed and evaluated using two groups, and the dataset is taken from the Kaggle Group 1 comprises 20 samples employing the ESVM methodology, while Group 2 consists of 20 samples applying the Lasso Regression methodology, Utilizing a G Power calculator with a pre-test power of 80% and an alpha value of 0.05 obtained. ESVM algorithm achieved an accuracy of 86.50%, while the Lasso Regression algorithm achieved an accuracy of 84.51%. The independent results reveal that ESVM method is significantly better than the Lasso regression technique for analysing and identifying facial age, with an 86.50 % accuracy rate.