| Autor: |
Henao, Daniel, Lopez, Jorge, Tobon, Jorge, Barrero, Cesar |
| Předmět: |
|
| Zdroj: |
Hyperfine Interactions; 4/25/2023, Vol. 244 Issue 1, p1-10, 10p |
| Abstrakt: |
Artificial intelligence (AI) is widely used for analyzing large data systems. Artificial neural networks (ANN) are a specialized AI method used for fitting and classifying patterns in data sets. This work presents the use of an artificial neural network as a recognition system in Mössbauer spectroscopy. Using a Mössbauer spectra database (Dyar, DEVAS Database, 2022) and generated spectra, the ANN was trained to identify whether the shape of a Mössbauer spectrum corresponds to hematite (α -Fe 2 O 3 ), magnetite (Fe 3 O 4 ), or neither. A precision percentage of 99% on the training data set and 100% on the validation set were achieved. An accuracy of 97% on the test set was achieved. These results confirm the efficiency of the ANN for identifying these compounds. The proposed method can be easily extended to identify other iron-containing compounds and for use with other characterization techniques. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
Full text from SpringerLink