Classification model for reducing absenteeism of nurses at hospitals using machine learning and artificial neural network techniques

Efficient production is paramount for all types of institutions, hinging upon the attainment of predefined employee targets and their subsequent outcomes. In contemporary times, a pervasive issue plaguing institutions is declining production, primarily stemming from employee absenteeism due to various reasons, ultimately eroding profitability. In our research, we spotlight organizations that rely heavily on healthcare services to bolster their bottom line, focusing on the unique case of King Abdullah University Hospital (KAUH). Drawing insights from surveys administered to nurses, we meticulously compiled a clean dataset using the OpenRefine tool. Subsequently, we harnessed the power of Machine Learning (ML) and Artificial Neural Network (ANN) techniques to construct a classification model. We judiciously assessed performance metrics such as Accuracy, Precision, and Recall to discern the most effective model. Our comparative analysis unequivocally underscored the superiority of ANN in our classification task, boasting an impressive 82% accuracy rate for predicting nurse absenteeism. This research endeavors to forecast the likelihood of nurse absenteeism in the upcoming year and elucidate the key contributing factors that elevate the risk of such occurrences. The overarching objective is to equip King Abdullah University Hospital’s decision-makers with a valuable tool to proactively mitigate absenteeism rates, elevate healthcare service quality, and elevate production levels, ultimately yielding optimal profitability.