Human Resource Management System based on Modified Genetic Algorithm

For the past few years, Employee attrition has been increasing rapidly which occurs when a worker leaves the company or an organization for unpredictable reasons. Also, Employee attrition leads to a major loss to the organization in many ways such as the pause of continuing tasks, risk of leaking core technology, re-training which takes extra time, cost of employees, and many more. Moreover, a human resource management system is a common problem often faced by Human Resource (HR) managers of the working organization or a company and the existing models like stacking ensembles have more overfltting problems. This study proposes a method namely Modified Genetic Algorithm (MGA) to address existing limitations for predicting employer attrition and improvement. Initially, the input of the International Business Machines Corporation – Human Resource (IBM-HR) dataset is fed into the Z-Score normalization technique that converts the data into the scalar format and performs preprocessing. Then, the preprocessed data is further processed into feature selection using MGA for selecting the most relevant features to predict employee turnover. Further, the processed MGA model is then classified based on the Naïve Bayes classifier (NB). MGA resulted in greater efficiency when compared with the existing methods such as stacking ensemble and Linear Support Vector Machine (L-SVM) in terms of accuracy (0.978), precision (0.999), f1-Score (0.978), and recall (0.974) respectively.