Conceptual Review of Dual-Target Machine Learning Models for Prediction of Reproductive Health in Women

Abstract

nfertility and menopause are two major interconnected challenges in women reproductive health. They are linked through ovarian ageing, endocrine changes and the progressive reduction in reproductive capacity, although infertility may also arise from causes unrelated to menopausal transition. This study is conceptually grounded in the application of dual-target machine learning models to improve the prediction of infertility risk and menopause transition. Most existing systems are singletarget models that focus on predicting one outcome at a time. They predicted either infertility or menopause. This limits their ability to capture shared biological patterns and reduces the potential for integrated reproductive health analysis. The dual-target machine learning approach addresses this gap by predicting infertility risk and menopause stage simultaneously from shared reproductive health data. This enables the model to learn common patterns associated with reproductive ageing and may improve predictive efficiency and clinical relevance. The study highlights the importance of dual-target modelling in supporting early risk identification, improving clinical decision-making and enhancing more personalised reproductive healthcare. It is particularly relevant in resource-limited settings where delayed diagnosis and fragmented health data are common. It advances the development of integrated reproductive intelligent systems for improved women's health outcomes.