A Machine Learning Approach for Employee Retention Prediction

Massive investment in employee skills training has been adopted by lots of organizations in reaction to the rapid evolution of the global trends and technology adoption. Unfortunately, target employee retention after training unsatisfactorily gives a negative return on investment. Prediction of target candidate decision before training and understanding the features that affect the candidate decision can greatly contribute to candidate selection and decision feature optimization process for increased employee retention. The method proposed in this paper successfully models and analyses various machine learning classifiers for illustrating features that affect the target candidate decision and predict the probability of candidate retention before training. Classical metrics are used to express the results of the algorithms used and the Random Forest Classifier revealed the finest percentage in accuracy summarized as 99.1%, 84.6%, 91.8% on the training, testing and overall dataset respectively.