Viscosity prediction of refining slag based on machine learning with domain knowledge

Key Points

• The viscosity of refining slag can be effectively predicted using machine learning techniques—improving process efficiency in metallurgical operations.
• The predictive model demonstrates a notable prediction accuracy of over 85% based on training datasets and domain knowledge.
• Analysis incorporates several parameters related to slag composition and processing conditions for accurate viscosity estimates.
• Enhanced models may lead to improved industrial applications, optimizing slag refinement processes for better material properties.