Likelihood-Based Machine Learning and Hierarchical Modeling for Clustered, Correlated, and Biomedical Data

The growing availability of clustered, longitudinal, multilevel, and high-dimensional data across biomedical, pharmaceutical, and health systems research has created both unprecedented opportunities and significant methodological challenges. While modern machine learning (ML) methods have achieved remarkable predictive success, their integration with principled likelihood-based statistical inference and hierarchical modeling remains an area of active and necessary development. This Research Topic seeks to bridge rigorous statistical foundations with modern machine learning methodologies, particularly for dependent, structured, and correlated data. Emphasis is placed on unifying likelihood-based inference, hierarchical modeling, uncertainty quantification, and responsible artificial intelligence (AI) to support robust, interpretable, and reproducible scientific discovery. We invite contributions that advance foundational and applied aspects of statistical methodology and machine learning, including but not limited to: - Statistical learning theory and generalization guarantees for modern AI systems - Bayesian modeling, hierarchical inference, and uncertainty quantification - High-dimensional inference (p ≫ n), regularization, and sparsity-inducing techniques - Likelihood-based machine learning frameworks - Model selection and probabilistic modeling principles - Kernel methods, nonparametric regression, and ensemble learning - Causal inference and probabilistic graphical models, particularly in health research - Interpretable and explainable machine learning with statistical guarantees - Responsible AI, bias detection, fairness, and robustness in predictive modeling - Federated learning and privacy-preserving statistical methodologies - Machine learning approaches tailored for biomedical, pharmaceutical, and clinical datasets - Applications in drug discovery, precision medicine, health systems analytics, epidemiological forecasting, and related domains Submissions may include theoretical developments, methodological innovations, computational advances, simulation studies, and applied case studies. By strengthening the mathematical and statistical foundations of modern AI systems, this Topic aims to promote principled, transparent, and reliable machine learning in high-impact scientific and societal applications. Topic editors: Edmund Ameyaw, Ernest Fokoué Source: https://www.frontiersin.org/research-topics/79119/likelihood-based-machine-learning-and-hierarchical-modeling-for-clustered-correlated-and-biomedical-data