Predictive Models for Industrial Applications

This article is a review that focuses on predictive models utilized in various industrial applications. It categorizes these models into three types: full-knowledge, zero-knowledge, and partial-knowledge based models. Full-knowledge models are characterized by high explainability and data efficiency but are computationally demanding during prediction. Zero-knowledge models, while highly accurate, are less explainable and require significant data. Partial-knowledge models aim to combine the strengths of both approaches. The review highlights key industrial sectors such as extraction, chemical processes, manufacturing, transportation, energy, and construction, while intentionally excluding sectors like health and agriculture for focused analysis. The authors conducted a meta-review to identify gaps in existing literature and provide a formal analysis of the subject. They also include illustrative examples and discuss unresolved challenges and future research directions, making this work a valuable resource for researchers and practitioners in the field.