Modeling REE interactions by Explainable Als
This study aims to develop a new explainable artificial intelligence (XAI) model and machine learning (ML) algorithms to integrate Rare Earth Elements (REEs) characterization data of different ores into a digitalized platform, which can potentially be used to predict the REE content and their chemical interactions with other elements within the raw material structures. The main goal of this study is to generate and develop a new XAI and a predictive ML model to enhance a robust platform for understanding the magnitude and significance of complex intercorrelations, educating lab operators, reducing potential environmental issues, and assisting sustainable critical raw material production. Such an AI platform can be used for various applications.
We are examining various AI models to explore the interactions and predict REE content based on other high concentration elements. A new XAI and a predictive ML model will be developed to enhance the modeling accuracy and will be compared with existing models. The potential outcome would be an AI model showing the interactions within various elements of REE resources that can be used to predict REE content.
Participants: Dr. Saeed Chehreh Chelgani (LTU Mineral Processing group & WISE) Dr. Elisa H Barney Smith (LTU ML) Prakash (LTU ML) Richa (LTU ML)
Funding: This work is partially funded by the Wallenberg AI, Autonomous Systems and Software Program (WASP) funded by the Knut and Alice Wallenberg Foundation
Updated: