Details

Neural Network-Based Classification of XRF Profiles Of Pottery Shards Using Synthetic Data

Ankita Nandy

Independent Researcher, Hooghly, West Bengal, India

65-71

Vol: 13, Issue: 3, 2023

Receiving Date: 2023-06-17 Acceptance Date:

2023-08-20

Publication Date:

2023-08-20

Download PDF

http://doi.org/10.37648/ijrst.v13i03.007

Abstract

Across all archaeological sites, pottery fragments have been abundant and insightful resources for understanding the societies which crafted, traded and/or used them. Their chemical profiles generated using X-ray fluorescence (XRF) techniques are used to characterize the raw materials and manufacturing techniques, and thus map them to their potential origins. Numerous researchers have published their XRF analysis results for pottery shards across the world, which exist as isolated small datasets. The collation of these datasets and subsequent usage in training classifiers aids in studying potential migration routes and trade and diplomatic relations which have bridged civilizations. Artificial Neural Networks (ANNs) have gained popularity across all domains considering their versatility in learning non-linear, complex patterns. However, training them requires large datasets which is less common in the field of archaeology. To solve the problem posed by data availability, the classifier ANN is trained with synthetically generated data. The 297 XRF records, when added to synthetically generated data, swelled to over 13k records, and could classify pottery shards across 11 geographies with an accuracy of over 80 per cent. The discounted online store provides wholesale rolex replica watches UK with Swiss movements.
Place an order for UK cheap replica rolex watches online is convenient and guaranteed.

Keywords: pottery shards; XRF; synthetic data; classification; neural networks

References

  1. Alqahtani, H., Kavakli-Thorne, M., & Kumar, G. (2021). Applications of generative adversarial networks (GANs): An updated review. Archives of Computational Methods in Engineering, 28, 525-552.
  2. Bickler, S. H. (2021). Machine learning arrives in archaeology. Advances in Archaeological Practice, 9(2), 186-191.
  3. Chandrasekaran, A., Naseerutheen, A., & Ravisankar, R. (2017). Dataset on elemental concentration and group identification of ancient potteries from Tamil Nadu, India. Data in brief, 10, 215-220.
  4. Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE: synthetic minority over-sampling technique. Journal of artificial intelligence research, 16, 321-357.
  5. Dablain, D., Krawczyk, B., & Chawla, N. V. (2022). DeepSMOTE: Fusing deep learning and SMOTE for imbalanced data. IEEE Transactions on Neural Networks and Learning Systems.
  6. Daszkiewicz, M., Gavrylyuk, N., Hellström, K., Kaiser, E., Kashuba, M., Kulkova, M., ... & Winger, K. (2020). Possibilities and limitations of pXRF as a tool for analysing ancient pottery: a case study of Late Bronze and Early Iron Age pottery (1100–600 BC) from the northern Black Sea region. Praehistorische Zeitschrift, 95(1), 238-266.
  7. Długosz-Lisiecka, M., Sikora, J., Krystek, M., Płaza, D., & Kittel, P. (2022). Novel method of ancient pottery analysis based on radioactive isotope ratios: a pilot study. Heritage Science, 10(1), 1-18.
  8. Elbashir Siddig, F., Elbashir, A. A., Lepper, V., & Hussein, A. (2018). Spectroscopic approach for characterization of archaeological potsherds excavated from some Neolithic sites from Sudan. International journal of experimental spectroscopic techniques, 3(2), 1-11.
  9. Figueira, A., & Vaz, B. (2022). Survey on synthetic data generation, evaluation methods and GANs. Mathematics, 10(15), 2733.
  10. Fusaro, A., Martínez Ferreras, V., Gurt Esparraguera, J. M., Angourakis, A., Pidaev, S. R., & Baratova, L. (2019). Islamic pottery from ancient Termez (Uzbekistan): new archaeological and archaeometric data. ArcheoSciences. Revue d'archéométrie, (43-2), 249-264.
  11. He, Z., Zhang, M., & Zhang, H. (2016). Data-driven research on chemical features of Jingdezhen and Longquan celadon by energy dispersive X-ray fluorescence. Ceramics International, 42(4), 5123-5129.
  12. Heyburn, R., Bond, R. R., Black, M., Mulvenna, M., Wallace, J., Rankin, D., & Cleland, B. (2018). Machine learning using synthetic and real data: similarity of evaluation metrics for different healthcare datasets and for different algorithms. In Data Science and Knowledge Engineering for Sensing Decision Support: Proceedings of the 13th International FLINS Conference (FLINS 2018) (pp. 1281-1291).
  13. Jones, C., Daly, N. S., Higgitt, C., & Rodrigues, M. R. (2022). Neural network-based classification of X-ray fluorescence spectra of artists’ pigments: an approach leveraging a synthetic dataset created using the fundamental parameters method. Heritage Science, 10(1), 1-14.
  14. Moradi, H., Sarhaddi-Dadian, H., Ramli, Z. & Rahman, N. H. S. N.A. (2013). Compositional analysis of the pottery shards of Shahr-I Sokhta, South Eastern Iran. Research Journal of Applied Sciences, Engineering and Technology, 6(4), 654-659.
  15. Oliveira, L. S. S., Abreu, C. M., Ferreira, F. C. L., Lopes, R. C. A., Almeida, F. O., Tamanaha, E. K., ... & Souza, D. N. (2020). Archeometric study of pottery shards from Conjunto Vilas and São João, Amazon. Radiation Physics and Chemistry, 167, 108303.
  16. Panda, S. S., Jena, G. N., & Garnayak, D. B. (2019). Characterization of Representative Ancient Potteries: Chemical, Mineralogical and Morphological Studies. International Journal of Conservation Science, 10(2), 317-326.
  17. Patki, N., Wedge, R., & Veeramachaneni, K. (2016). The synthetic data vault. In 2016 IEEE International Conference on Data Science and Advanced Analytics (DSAA) (pp. 399-410). IEEE.
  18. Ramli, Z., Rahman, N. H. S. N. A., Samian, A. L., Razman, M. R., Zakaria, S. Z. S., Jusoh, A., ... & Dadian, H. S. (2014). X-Ray Diffraction (XRD) and X-Ray Fluorescence (XRF) analysis of proto-historic votive tablets from Chawas cave, Hulu Kelantan, Malaysia. Research Journal of Applied Sciences, Engineering and Technology, 7(7), 1381-1387.
  19. Reddy, A., Attaelmanan, A. G., & Mouton, M. (2012). Pots, plates and provenance: sourcing Indian coarse wares from Mleiha using X-ray fluorescence (XRF) spectrometry analysis. In IOP Conference Series: Materials Science and Engineering (Vol. 37, No. 1, p. 012010). IOP Publishing.
  20. Sarhaddi-Dadian, H., Ramli, Z., Rahman, A., & Mehrafarin, R. (2015). X-ray diffraction and X-ray fluorescence analysis of pottery shards from new archaeological survey in south region of Sistan, Iran. Mediterranean Archaeology and Archaeometry, 15(3), 45-56.
  21. Sarhaddi-Dadian, H., Moradi, H., Zuliskandar, R., & Purzarghan, V. (2017). X-ray fluorescence analysis of the pottery shards from dahan-E ghulaman, the achaemenid site in Sistan, east of Iran. Interdisciplinaria Archaeologica, 8(1), 35-41.
  22. Smith, R. (2007, September). An overview of the Tesseract OCR engine. In Ninth international conference on document analysis and recognition (ICDAR 2007) (Vol. 2, pp. 629-633). IEEE.
  23. Wu, Q. Q., Zhu, J. J., Liu, M. T., Zhou, Z., An, Z., Huang, W., ... & Zhao, D. Y. (2013). PIXE-RBS analysis on potteries unearthed from Lijiaba Site. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 296, 1-6.
  24. Xu, L., Skoularidou, M., Cuesta-Infante, A., & Veeramachaneni, K. (2019). Modeling tabular data using conditional gan. Advances in neural information processing systems, 32.
  25. Zhao, Z., Kunar, A., Birke, R., & Chen, L. Y. (2021, November). Ctab-gan: Effective table data synthesizing. In Asian Conference on Machine Learning (pp. 97-112). PMLR.
Back

Disclaimer: All papers published in IJRST will be indexed on Google Search Engine as per their policy.

We are one of the best in the field of watches and we take care of the needs of our customers and produce replica watches of very good quality as per their demands.