Enhanced honey badger algorithm and random forest based mechanism for iris detection in authenticate users in information systems

Abstract

Iris recognition became the important element in new authentication systems because of its high accuracy and robustness. In this paper, we present the developed Honey Badger Algorithm (HBA) combined with classification algorithm based on Random Forest for iris diagnosis in info systems. Presented development defines Levy Flight algorithm for developing global search abilities, adaptive inertia weight to balance local and global searches, active weighting to set exploration and exploitation trade-off in runtime. The generally accessible iris set of data is used for validating presented technique. Developed HBA performance is compared to traditional HBA and other state-of-the-art optimization methods. Assessment is performed using several metrics, such as accuracy and Error Rate. Experimental outcomes illustrate that presented developed HBA considerably develops classification accuracy and feature selection efficiency, making it satisfactory strategy for safe and reliable user authentication in info systems. Additionally, the proposed algorithm is compared with other methods, showing that the proposed method has higher classification accuracy (97.44%) and lower Classification Error Rate (2.56%).