Contribution to a conference proceedings/Contribution to a book

IMPULSE-2025-00003

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Automation of PGAA Spectra Analysis with Deep Learning

Boschmann, D. (Corresponding author) ; Stieghorst, C.MLZ* ; Knezevic, D.MLZ* ; Kadri, L.MLZ* ; Niggemann, O.

2024
IEEE

Please use a persistent id in citations: doi:10.1109/INDIN58382.2024.10774320

Abstract: Analyzing Prompt Gamma Activation Analysis (PGAA) spectra poses significant challenges, particularly in Traditional expert analysis, while effective, is time-consuming. This paper addresses the need for an efficient, automated solution to enhance the analysis process. The research investigates the use of machine learning (ML) and deep learning (DL) algorithms in the automated analysis of PGAA spectra. We aim to establish a new metric for comparing automated analysis with expert analysis, providing a baseline using Linear Regression, Random Forest, 1D Convolutional Neural Network, Feed Forward Neural Network, and Autoencoder algorithms. Established metrics like Mean Square Error (MSE) and Mean Absolute Error (MAE) are utilized to compare the performance of these automated approaches against traditional expert analysis. Using actual spectra from various research projects and semireal augmented data, the study demonstrates that the Feed Forward Neural Network (FFNN) and Autoencoder algorithms can effectively predict the magnitude of the present elements. These findings suggest especially DL algorithms could significantly assist researchers and industry personnel by providing a rough overview of the material and saving valuable time. However, the automated approach requires further refinement, particularly in handling noisy data, predicting additional crucial information, and integrating more prior knowledge into the analysis. This research offers valuable insights into the application of ML algorithms in spectral analysis and lays a foundation for further advancements in the field.

Keyword(s): Instrument and Method Development (1st) ; Instrument and Method Development (2nd)

---

Contributing Institute(s):

1. PGAA (PGAA)
2. DEVA (DEVA)

Research Program(s):

1. 05D23WO2 - Verbundprojekt 05D2022 - EvalSpek-ML: Entwicklung und Evaluierung von Machine-Learning-Algorithmen für die Analyse konvexkombinierter spektraler Daten. Teilprojekt 4. (BMBF-05D23WO2) (BMBF-05D23WO2)

Experiment(s):

1. PGAA: Prompt gamma activation analysis (NL4b)

---