Enhanced Ethnicity Prediction from Iris Images Using PCA-SVM Optimized with Genetic Algorithm

Biometric recognition has gained significant importance in recent years due to its potential applications in security, forensics, and demographic classification. Among various biometric modalities, iris recognition has emerged as one of the most reliable because of its unique texture and stability throughout a person’s lifetime. This paper presents novel research for an ethnicity prediction system for Yoruba sub-populations using iris images with a hybrid approach that combines Principal Component Analysis (PCA) for feature reduction, Support Vector Machine (SVM) for classification, and Genetic Algorithm (GA) for optimization. The proposed PCA-SVM-GA system was evaluated on benchmark iris datasets to investigate its effectiveness in demographic classification. The SVM-GA achieved better performance than SVM across metrics like accuracy at 87.71% vs. 83.75% and specificity at 91.88% vs. 88.75%, also an optimal threshold of 0.75. The experimental results demonstrate that the hybridized model outperforms conventional PCA-SVM methods, achieving higher prediction accuracy, robustness, and generalization. Hence, SVM-GA's potential in ethnicity prediction enhances biometric system applications. Keywords: Iris Biometrics, Intra-ethnic Classification, Genetic Algorithm, Yoruba Ethnicity, PCA and SVM