Predicting non-violent work behaviour among employees using machine learning techniques

Abstract

Purpose This study aims to develop a model to predict non-violent work behaviour (NVWB) among employees using machine learning techniques. Design/methodology/approach Four machine learning techniques (Naïve Bayes, decision tree, logistic regression and ensemble learning) were used to develop a prediction model for NVWB of employees. Also, 10-fold cross-validation method was used to validate the NVWB prediction models. The confusion matrix is used to derive various performance matrices to express the predictive capability of NVWB models quantitatively. Findings The model developed using random forest technique was identified as best NVWB prediction model, as it resulted in highest true positive rate and true negative rate, thereby resulting in the highest geometric mean, balance and area under receiver operator characteristics curve. Originality/value To the best of the authors’ knowledge, this is one of the pioneer studies that used machine learning techniques to develop a predictive model of NVBW.