Journal of Artificial Intelligence and Metaheuristics
Journal DOI
2833-5597ISSN (Online)
Volume 11 , Issue 1 , PP: 117-146, 2026 | Cite this article as | XML | Html | PDF | Full Length Article
Ebrahim A. Mattar 1 * , S. K. Towfek 2
Doi: https://doi.org/10.54216/JAIM.110105
Gamma–hadron discrimination remains a fundamental challenge in very-high-energy gamma-ray astronomy due to the strong overlap between gamma-ray–initiated and hadron-induced air showers recorded by imaging atmospheric Cherenkov telescopes, particularly at low energies where background contamination is severe. Traditional cut-based and non-optimized machine learning approaches often struggle to fully exploit the nonlinear and correlated nature of Cherenkov image parameters, leading to suboptimal background suppression and reduced telescope sensitivity. To address these limitations, this paper proposes a unified deep learning and metaheuristic optimization framework that combines an enhanced attention-based long short-term memory network (EALSTM) with advanced optimization strategies. In particular, a novel Adaptive Balanced Greylag Goose Optimization algorithm (ABGGO) is employed to jointly perform feature selection and hyperparameter optimization, enabling effectiveexploration–exploitation balancing while preserving physically meaningful feature representations. The proposed ABGGO+EALSTM framework is systematically evaluated against baseline deep learning models, including artificial neural networks (ANN), convolutional neural networks (CNN), and standard long short-term memory networks (LSTM), under identical experimental conditions. Experimental results on a Monte Carlo–generated Cherenkov telescope dataset demonstrate clear and consistent performance gains at every stage of the analysis. In the baseline evaluation stage, EALSTM achieves an accuracy of 0.9294 and an F-score of 0.9266, outperforming ANN, CNN, and LSTM. Following metaheuristic optimization, the proposed ABGGO+EALSTM model attains a peak accuracy of 0.9718, sensitivity of 0.9694, specificity o f 0 .9740, a nd F-score o f 0 .9705, representing absolute improvements exceeding 4% over the baseline EALSTM configuration and outperforming GA+EALSTM, GWO+EALSTM, and PSO+EALSTM variants. These results demonstrate that integrating attention-based deep learning with adaptive metaheuristic optimization significantly enhances gamma–hadron discrimination, leading to improved background suppression and signal retention. The proposed framework offers a scalable and robust solution for current and next-generation Cherenkov observatories, with strong potential for real-time event filtering, multi-telescope analysis, and future deployment on real observational data.
Gamma&ndash , hadron discrimination , Imaging atmospheric Cherenkov telescopes , Attention-based deep learning , Metaheuristic optimization , Greylag Goose Optimization
[1] J. Biteau and M. Meyer, “Gamma-ray cosmology and tests of fundamental physics,” Galaxies, vol. 10, no. 2, 2022, issn: 2075-4434. doi: 10.3390/galaxies10020039. [Online]. Available: https://www.mdpi.com/2075-4434/10/2/39.
[2] F. Giovannelli, “Gamma-ray bursts: The energy monsters of the universe,” Galaxies, vol. 13, no. 2, 2025, issn: 2075-4434. doi: 10.3390/galaxies13020016. [Online]. Available: https: //www.mdpi.com/2075-4434/13/2/16.
[3] F. Casaburo, S. Ciprini, D. Gasparrini, and F. Giacchino, “Sixteen years of gamma-ray discoveries and agn observations with fermi-lat,” Particles, vol. 8, no. 1, 2025, issn: 2571-712X. doi: 10.3390/particles8010017. [Online]. Available: https://www.mdpi.com/2571-712X/8/1/17.
[4] L. A. Stuani Pereira and S. V. Bernardo da Silva, “From gamma rays to cosmic rays: Lepto-hadronic modeling of blazar sources as candidates for ultra-high-energy cosmic rays,” Universe, vol. 11, no. 8, 2025, issn: 2218-1997. doi: 10.3390/universe11080266. [Online]. Available: https://www.mdpi.com/2218-1997/11/8/266.
[5] X. Yu et al., “Establishment of a breeding approach combined with gamma ray irradiation and tissue regeneration for highbush blueberry,” Agronomy, vol. 15, no. 1, 2025, issn: 2073-4395. doi: 10.3390/agronomy15010217. [Online]. Available: https://www.mdpi.com/2073-4395/ 15/1/217.
[6] Y. Sato, K. Murase, Y. Ohira, S. Inoue, and R. Yamazaki, “Two-component jet model for the afterglow emission of grb 201216c and grb 221009a and implications for jet structure of very-high-energy gamma-ray bursts,” Journal of High Energy Astrophysics, vol. 48, p. 100 415, 2025, issn: 2214-4048. doi: https://doi.org/10.1016/j.jheap.2025.100415. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S2214404825000965.
[7] R. Mirzoyan, “Technological novelties of ground-based very high energy gamma-ray astrophysics with the imaging atmospheric cherenkov telescopes,” Universe, vol. 8, no. 4, 2022, issn: 2218-1997. doi: 10.3390/universe8040219. [Online]. Available: https://www.mdpi.com/2218-1997/8/ 4/219.
[8] G. D’Amico, “Statistical tools for imaging atmospheric cherenkov telescopes,” Universe, vol. 8, no. 2, 2022, issn: 2218-1997. doi: 10.3390/universe8020090. [Online]. Available: https: //www.mdpi.com/2218-1997/8/2/90.
[9] J. Sitarek, M. Pecimotika, N. Żywucka, D. Sobczyńska, A. Moralejo, and D. Hrupec, “Estimation of the atmospheric absorption profile with isotropic background events observed by imaging atmospheric cherenkov telescopes,” Journal of High Energy Astrophysics, vol. 42, pp. 87–95, 2024, issn: 2214-4048. doi: https://doi.org/10.1016/j.jheap.2024.03.003. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S221440482400020X.
[10] D. Dominis Prester et al., “Characterisation of the atmosphere in very high energy gamma-astronomy for imaging atmospheric cherenkov telescopes,” Universe, vol. 10, no. 9, 2024, issn: 2218-1997. doi: 10.3390/universe10090349. [Online]. Available: https://www. mdpi.com/2218-1997/10/9/349.
[11] E. R. Owen, K. Wu, Y. Inoue, H.-Y. K. Yang, and A. M. W. Mitchell, “Cosmic ray processes in galactic ecosystems,” Galaxies, vol. 11, no. 4, 2023, issn: 2075-4434. doi: 10 . 3390 / galaxies11040086. [Online]. Available: https://www.mdpi.com/2075-4434/11/4/86.
[12] L. Tibaldo, D. Gaggero, and P. Martin, “Gamma rays as probes of cosmic-ray propagation and interactions in galaxies,” Universe, vol. 7, no. 5, 2021, issn: 2218-1997. doi: 10.3390/ universe7050141. [Online]. Available: https://www.mdpi.com/2218-1997/7/5/141.
[13] J. Balibrea-Correa et al., “Hadron therapy range verification via machine-learning aided prompt-gamma imaging,” in 2021 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 2021, pp. 1–7. doi: 10.1109/NSS/MIC44867.2021.9875637.
[14] M. Zehtabvar, K. Taghandiki, N. Madani, D. Sardari, and B. Bashiri, “A review on the application of machine learning in gamma spectroscopy: Challenges and opportunities,” Spectroscopy Journal, vol. 2, no. 3, pp. 123–144, 2024, issn: 2813-446X. doi: 10.3390/spectroscj2030008. [Online]. Available: https://www.mdpi.com/2813-446X/2/3/8.
[15] M. Kupczynski, “Mathematical modeling of physical reality: From numbers to fractals, quantummechanics and the standard model,” Entropy, vol. 26, no. 11, 2024, issn: 1099-4300. doi: 10.3390/e26110991. [Online]. Available: https://www.mdpi.com/1099-4300/26/11/991.
[16] M. H. L. d. Boni, I. Gervasio Sene Junior, and R. Martins da Costa, “Tabular data augmentation using artificial intelligence: A systematic review and taxonomic framework,” IEEE Access, vol. 13, pp. 138 950–138 969, 2025. doi: 10.1109/ACCESS.2025.3593449.
[17] K. Karthick, S. A. Agnes, S. S. Kumar, S. Alfarhood, and M. Safran, “Identification of high energy gamma particles from the cherenkov gamma telescope data using a deep learning approach,” IEEE Access, vol. 12, pp. 16 741–16 752, 2024. doi: 10.1109/ACCESS.2024.3359533.
[18] S. Truzzi, G. Bonnoli, R. Cappuccio, R. Paoletti, and T. Miener, “Events classification in magic through convolutional neural network trained with images of observed gamma-ray events,” in Machine Learning for Astrophysics, F. Bufano, S. Riggi, E. Sciacca, and F. Schilliro, Eds., Cham: Springer International Publishing, 2023, pp. 187–191, isbn: 978-3-031-34167-0.
[19] A. Brill, Q. Feng, T. B. Humensky, B. Kim, D. Nieto, and T. Miener, “Investigating a deep learning method to analyze images from multiple gamma-ray telescopes,” in 2019 New York Scientific Data Summit (NYSDS), 2019, pp. 1–4. doi: 10.1109/NYSDS.2019.8909697.
[20] A.-Y. Cheng et al., “Application of deep learning methods combined with physical background in wide field of view imaging atmospheric cherenkov telescopes,” Nuclear Science and Techniques, vol. 35, no. 4, p. 84, 2024. doi: https://doi.org/10.1007/s41365-024-01448-8.
[21] A. Pathania, K. Singh, S. Singh, A. Tolamatti, B. Singh, and K. Yadav, “Identification of gamma ray pulsar candidates in the fermi-lat 4fgl-dr4 unassociated sources using supervised machine learning,” Astroparticle Physics, vol. 175, p. 103 185, 2026, issn: 0927-6505. doi: https://doi.org/10.1016/j.astropartphys.2025.103185. [Online]. Available: https: //www.sciencedirect.com/science/article/pii/S0927650525001082.
[22] M. P. Das, V. Dhar, A. Singh, N Bhatt, and K. Yadav, “Development of source-dependent gamma-ray image parameters using generative adversarial network (gan) for the mace telescope and its implications,” Journal of High Energy Astrophysics, p. 100 416, 2025. doi: https : //doi.org/10.1016/j.jheap.2025.100416.
[23] H. Mehta, S. Singh, V. Pandey, P. Verma, and A. Antony, “Autonomous telescopes and observatories,” The Intelligent Universe: AI’s Role in Astronomy, pp. 313–357, 2025.
[24] F. Farsian et al., “Benchmarking quantum convolutional neural networks for signal classificationin simulated gamma-ray burst detection,” in 2025 33rd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP), 2025, pp. 372–380. doi: 10.1109/ PDP66500.2025.00059.
[25] A. Adelfio et al., “Anomaly detection with machine learning on time series: Unveiling lost transients data,” in 2025 33rd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP), 2025, pp. 396–403. doi: 10.1109/PDP66500.2025.00062.
[26] E. Gres et al., “Method for determination of gamma-ray energy from taiga-iact data based on analysis of autoencoder-derived essential features,” Physics of Atomic Nuclei, pp. 1–6, 2025. doi: https://doi.org/10.1134/S1063778825090182.
[27] B. Khek, A. Mishra, M. Buuck, and T. Shutt, “Gamma ray source localization for time projection chamber telescopes using convolutional neural networks,” AI, vol. 3, no. 4, pp. 975–989, 2022,issn: 2673-2688. doi: 10.3390/ai3040058. [Online]. Available: https://www.mdpi.com/2673- 2688/3/4/58.
[28] M. Das, V. Dhar, and K. Yadav, “Generative adversarial network based method for generation of synthetic image parameters for tactic γ-ray telescope,” Astronomy and Computing, vol. 44, p. 100 741, 2023. doi: https://doi.org/10.1016/j.ascom.2023.100741.
[29] Y. Htet, M. Sudvarg, A. Butzel, J. D. Buhler, R. D. Chamberlain, and J. H. Buckley, “Machine learning aboard the adapt gamma-ray telescope,” in SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, 2024, pp. 4–10. doi: 10.1109/SCW63240.2024.00008.
[30] A. Sinha, M. Shahid, A. Nandan, C. Iwendi, A. K. Giri, and S. Anand, “A novel approach ofmachine learning application in astrophysics: Morphological feature wrapping based ensemble method for galaxy shape classification using gama dataset,” in Proceedings of ICACTCE’23 — The International Conference on Advances in Communication Technology and Computer Engineering, C. Iwendi, Z. Boulouard, and N. Kryvinska, Eds., Cham: Springer Nature Switzerland, 2023, pp. 593–603, isbn: 978-3-031-37164-6.
