Prediction modelling in career management

Abstract

A question that is often arose on career management is how to choose potential employees to become chief and achieve performance target based on employees' historical data. This research attempts to answer the question and tries to determine what factors can affect an employee to become a chief and capable to achieve performance target. To address the question, this study adopts predictive modelling approach using classification techniques and kl-divergence function. The research proposes some recommendations based on applied algorithm. Additionally, the study shows it is easier to choose a chief based on their historical data compare to predicting which employee can perform well in the future or predicting both potential chief and high performer employees. If organization wants to select potential employees to become chief and predicts them achieving their target, classification techniques should include all employees as the population and avoid to choose only subsets to ensure the confidence of result. Meanwhile, KL-Divergence function generates competencies that affect employee who become a chief and achieve the target. They are achievement orientation, adaptability, analytical thinking, communication and information seeking.