Biometric Verification Enhancement with Ensemble Learning Classifiers

Abstract

This research study asserts the potential of ensemble learning classifiers over traditional classification approaches in the field of biometric recognition. Three biometric traits: fingerprint, online signature and offline signature are considered to demonstrate the improvement in the accuracies using ensemble classifiers. Data standardization techniques are applied to scale the data in a standard form. Histogram of Oriented Gradients is applied to extract the features from the data and then fed to different classification techniques such as Support Vector Machine (SVM), k-nearest neighbors algorithm (k-NN), Multilayer Perceptron (MLP) Neural Network, Extra Tree Classifier (ETC) and Random Forest Classifier (RFC). It is evident from the experimental results acquired during this research that the ensemble learning classifiers outperformed the conventional classification methods.