International Conference on Data Science, Agents & Artificial Intelligence · Mar 26, 2026
XAI Based Early Menopause Detection using Super Learner and Discriminant Analysis
By Selvan MP, Kalaivani SD, Parida K et al.
The early detection of menopause is essential for ensuring women’s good health status. This paper presents an Explainable Artificial Intelligence-based machine learning approach for predicting the early onset of menopause using various types of data, including physiological, biochemical, psychological, lifestyle, and text information. The machine learning techniques we will apply include Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Random Forest (RF), and Super Learner (SL), because they account for complex relationships among features. The main biological and psychological characteristics taken into account in our research include weight fluctuations, anxiety disorders, irritability, insomnia, irregular menstrual periods, and hot flashes. Additionally, our study will cover biochemical markers including estrogen, follicle-stimulating hormone (FSH), and body mass index (BMI). Moreover, we will investigate the effects of smoking, drinking alcohol, and physical exercise. To select appropriate features, we will adopt the LASSO approach. To achieve model interpretability, the SHAP framework will be employed. SHAP stands for Shapley Additive Explanations, a method for interpreting machine learning model outputs based on feature contributions. We will evaluate the performance of our models in terms of accuracy, precision, recall, F1-score, and the Receiver Operating Characteristic (ROC) Area Under the Curve (AUC). The experimental results reveal the superiority of the Super Learner ensembles over single machine learning algorithms, thus illustrating the applicability of the ensemble learning approach for predicting early menopause onset. The experimental results show that the Explainable Artificial Intelligence approach is vital for making machine learning models more understandable and trustworthy for medical applications.