Optimizing models performance: a comprehensive approach to addressing employee turnover with artificial intelligence

Abstract

This study explores the application of three machine learning classification models—K-Nearest Neighbors (KNN), Logistic Regression (LR), and Gaussian Naive Bayes (GNB)—to predict employee attrition using a dataset from Kaggle. The dataset includes various employee attributes such as satisfaction level, last evaluation, number of projects, average monthly hours, time spent in the company, work accidents, promotions, department, salary, and attrition status. KNN utilizes a nonparametric approach to impute missing values and make predictions based on Euclidean distance. LR predicts the probability of employee attrition using regression coefficients, while GNB assumes normal distribution for continuous features. Model performance was evaluated using accuracy, confusion matrix, precision, recall, and F1-score. Hyperparameter optimization was conducted to enhance the models' predictive performance. The results demonstrate the effectiveness of these models in predicting employee attrition, with each offering unique advantages based on the data characteristics. This study provides valuable insights for organizations seeking to understand and reduce employee turnover through advanced machine learning techniques.