Explainable AI for Predictive Analytics on Employee Promotion

The success of a company depends upon the strategy, technology, finance, and more importantly, the performance and competence of its employees. To efficiently manage the manpower of an organization, the founders, CEO, and managers need to determine which employees must be promoted. Promotion boosts employee morale, improves the resource allocation of the organization, and helps identify talent. In this project, we have explored various machine learning algorithms to predict the promotion of an employee based on multiple employee-related factors such as education, training, rating and length of service. The analysis focuses on creating a predictive model after thorough feature analysis and preprocessing of the training data. We have trained multiple classification models such as logistic regression, decision tree, random forest, and XGBoost. Evaluation metrics, namely, precision, recall, and Fl-score, were used to evaluate the performance of each model, and the XGBoost classifier outperformed the other algorithms. In addition, we have implemented explainable artificial intelligence methods such as local interpretable model agnostic explanations. Additive explanations to derive the reason behind the predictions made by artificial intelligence. The current work managed to predict with an accuracy of 95% and detect the features that have an impact on employee promotion.