Enhancing employee churn prediction with weibull time-to-event modeling

A Very common problem of employee attrition that requires precise and fast predictive models in modern business settings. We proposed a novel time-to-event Weibull Time-to-Event Recurrent Neural Network forecasting model in this study. This is cross-verified and validated by conducting an in-depth comparative study regarding the performance of the proposed Weibull Time to Event Recurrent Neural Network model against traditional models used in machine learning. This involved the comparison of a number of the classical models on their ability and performance, noting the strengths and weaknesses of the Weibull Time to Event Recurrent Neural Network improvements realized with long-range dependencies. Application of the HR dataset had shown that, while using the Weibull Time to Event Recurrent Neural Network model, the prediction worked out well as compared to other models. Our model R-squared value is 0.98, with very minimal Mean Squared Error (MSE) of 0.0032. This model makes it easier for human resource (HR) professionals and decision-makers to understand the predictions that it makes, enabling them to develop recruitment and retention strategies better adjusted to the needs of given employees.