Machine Learning to Evaluate Important Human Capital (HC) Determinants Impacting IT Compensation

Abstract

India is producing over 1 million youth per month resulting in ameliorating economy and a high attrition rate in the workforce, therefore employers must leverage their benefits to perpetuate, nurture their rising workforce & avoid a shortfall in their talent pipeline to stay sustainable & be Employer of Choice. Corporates must develop a suitable compensation strategy to stay competitive and engage their workforce because employees depend mainly on wages and salaries which must be equivalent to their work done. In light of these, it is very much essential for the author to consciously evaluate human capital indicators that are impacting the compensation in Indian IT Companies. This paper aims to evaluate five different machine learning classifier algorithms to predict the topmost important features and the best model for IT compensation based on a survey of 1170 IT professional responses across 61 organizations collected in the NCR region. Due diligence has been given to evaluate the most accurate classifier based on the accuracy score. The result indicated that Random forest regression performs better with a mean absolute error of 0.07 degrees and an accuracy of 99.63%. The finding of the study reflects that the top important variables in determining CTC are Experience, Institution from which the candidate graduated, education, and the Skillset that the individual possesses. These variables have a greater impact on compensation designing and act as strong predictors while other variables are marginally insignificant for predicting the compensation of IT sector employees. Moreover, this study could be beneficial for job seekers and employers looking to hire top talent for their organizations. As the Deep Learning model requires a huge amount of data which was limited in this case, therefore, the researcher could not employ them.