Artificial Immune-based Algorithm for Academic Leadership Assessment

Abstract

Artificial Immune-based algorithm is inspired by the biological immune system as computational intelligence approach in data analysis. Negative selection algorithm is derived from immune-based algorithm’s family that used to recognize the pattern’s changes perform by the gene detectors in complementary state. Due to the self-recognition ability, this algorithm is widely used to recognize the abnormal data or non-self especially for fault diagnosis, pattern recognition, network security etc. In this study, the self-recognition performance proposed by the negative selection algorithm been considered as a potential technique in classifying employee’s competency. Assessing the employee’s performance in organization is an important task for human resource management people to identify the right candidate in job promotion assessment. Thus, this study attempts to propose an immune-based model in assessing academic leadership performance. There are three phases involved in experimental phase i.e. data acquisition and preparation; model development; and analysis and evaluation. The data consists of academic leadership proficiency was prepared as data-set for learning and detection processes. Several experiments were conducted using cross validation process on different model to identify the most accurate model. Therefore, the accuracy of NS classifier is considered acceptable enough for this academic leadership assessment case study. For enhancement, other immune-based algorithm or bio-inspired algorithms, such as genetic algorithm, particle swam optimization, ant colony optimization would also be considered as a potential algorithm for performance assessment.