EPFL scientists have developed a new AI-based technique to improve the chemical analysis of nanomaterials. Traditional methods often struggle with noisy data and overlapping signals, making it difficult to accurately identify and quantify different elements within a sample.
The team’s approach, published in the journal Nano Letters, is called PSNMF (non-negative matrix factorization-based pan-sharpening) and addresses these challenges by enhancing the clarity and accuracy of data from energy-dispersive X-ray spectroscopy (EDX). By combining machine learning techniques with advanced data processing, PSNMF allows for more precise analysis of nanomaterials, aiding research and development in fields like electronics and medicine.
