Guided Machine Learning

Guided machine learning (GML) integrates scientific knowledge with machine learning algorithms to improve model accuracy, interpretability, and generalizability. Current research emphasizes hybrid approaches combining data-driven models (like neural networks, including MLPs, CNNs, and RNNs) with established scientific principles or mechanistic models, focusing on enhancing predictive power and ensuring scientific consistency. This approach is particularly valuable in fields like chemical engineering and environmental science, where it facilitates more reliable process optimization, improved model explainability, and a deeper understanding of complex systems. The resulting models offer a powerful blend of data-driven insights and established scientific understanding.