Development of a Hybrid Classical-Quantum SVM Model for Predicting Perovskite Formation Energy
DOI:
https://doi.org/10.30871/jaic.v10i2.12240Keywords:
Formation Energy, Explainable AI, Hybrid SVM, Perovskite, Quantum Machine LearningAbstract
The global energy crisis and the threat of climate change are driving the acceleration of the transition towards new and renewable energy (NRE), with perovskite solar cells emerging as a leading candidate due to their high efficiency (>25%) and low production costs. A major challenge in the development of perovskite materials is their structural stability, which can be assessed through Formation Energy (FE). However, FE calculations using Density Functional Theory (DFT) are computationally expensive and not scalable for screening thousands of material candidates. This research develops a Quantum Machine Learning (QML) model to predict the Formation Energy of ABX₃ perovskite materials as an efficient alternative to conventional DFT approaches. Three variants of Support Vector Machine algorithms are compared comprehensively: the classical SVM with an RBF kernel as a baseline, Quantum SVM (QSVM) utilizing a quantum kernel with ZZFeatureMap to represent data in a high-dimensional quantum feature space, and Hybrid SVM (HSVM), which combines the expressiveness of the quantum feature map with the flexibility of classical parameter optimization. Evaluation results show that Hybrid SVM outperforms others with R² = 0.7221, RMSE = 0.4203, and MAE = 0.3228 on test data, improving by 4-8% compared to classical SVM (R² = 0.6674) and surpassing QSVM (R² = 0.6966). Interpretability analysis using Explainable AI with permutation feature importance reveals that the bandgap (HSE gap indirect) and effective hole mass are the most crucial predictors, confirming that electronic properties dominate the thermodynamic stability of perovskites. This study validates the potential of QML, particularly the hybrid approach, as an intelligent solution to accelerate the screening and design of renewable energy materials, supporting Indonesia's target of a 23% energy mix by 2025 and the Vision of Indonesia 2045.
Downloads
References
[1] Nunuk Febriananingsih, “Tata Kelola Energi Terbarukan di Sektor Ketenaglistrikan dalam Kerangka Pembangunan Hukum Nasional,” Majalah Hukum Nasional, vol. 49, no. 2, pp. 29–56, Nov. 2019, doi: 10.33331/mhn.v49i2.31.
[2] T. Tian et al., “Bandgap prediction of ABX3-type perovskites using Broad Learning System,” Mater. Today Commun., vol. 37, p. 107026, Dec. 2023, doi: 10.1016/j.mtcomm.2023.107026.
[3] M. Shafiq, B. Amin, M. A. Jehangir, A. R. Chaudhry, and G. Murataza, “First-principle calculations to investigate mechanical and acoustical properties of predicted stable halide Perovskite ABX3,” J. Mol. Graph. Model., vol. 133, p. 108861, Dec. 2024, doi: 10.1016/j.jmgm.2024.108861.
[4] H. Wang, Q. Zhang, M. Qiu, and B. Hu, “Synthesis and application of perovskite-based photocatalysts in environmental remediation: A review,” J. Mol. Liq., vol. 334, p. 116029, Jul. 2021, doi: 10.1016/j.molliq.2021.116029.
[5] M. Akrom, S. Rustad, and H. K. Dipojono, “A machine learning approach to predict the efficiency of corrosion inhibition by natural product-based organic inhibitors,” Phys. Scr., vol. 99, no. 3, p. 36006, Aug. 2024, doi: 10.1088/1402-4896/ad28a9.
[6] W. Hu and L. Zhang, “High-throughput calculation and machine learning of two-dimensional halide perovskite materials: Formation energy and band gap,” Mater. Today Commun., vol. 35, p. 105841, Jun. 2023, doi: 10.1016/j.mtcomm.2023.105841.
[7] S. Badini, S. Regondi, and R. Pugliese, “Unleashing the Power of Artificial Intelligence in Materials Design,” Materials, vol. 16, no. 17, p. 5927, Aug. 2023, doi: 10.3390/ma16175927.
[8] F. Emmert‐Streib and M. Dehmer, “Taxonomy of machine learning paradigms: A data‐centric perspective,” WIREs Data Mining and Knowledge Discovery, vol. 12, no. 5, Sep. 2022, doi: 10.1002/widm.1470.
[9] S. B. Akinpelu et al., “Interpretable machine learning methods to predict the mechanical properties of ABX3 perovskites,” Results Phys., vol. 65, p. 107978, Oct. 2024, doi: 10.1016/j.rinp.2024.107978.
[10] D. O. Obada, E. Okafor, S. A. Abolade, A. M. Ukpong, D. Dodoo-Arhin, and A. Akande, “Explainable machine learning for predicting the band gaps of ABX3 perovskites,” Mater. Sci. Semicond. Process., vol. 161, Jul. 2023, doi: 10.1016/j.mssp.2023.107427.
[11] A. Alhashmi, M. B. Kanoun, and S. Goumri-Said, “Machine Learning for Halide Perovskite Materials ABX3 (B = Pb, X = I, Br, Cl) Assessment of Structural Properties and Band Gap Engineering for Solar Energy,” Materials, vol. 16, no. 7, p. 2657, Mar. 2023, doi: 10.3390/ma16072657.
[12] Q. Tao, P. Xu, M. Li, and W. Lu, “Machine learning for perovskite materials design and discovery,” NPJ Comput. Mater., vol. 7, no. 1, p. 23, Jan. 2021, doi: 10.1038/s41524-021-00495-8.
[13] T. Zhang, “Bandgap prediction of MXene materials via multiple machine learning approaches,” J. Phys. Conf. Ser., vol. 3112, no. 1, p. 012093, Sep. 2025, doi: 10.1088/1742-6596/3112/1/012093.
[14] Z. Deng et al., “Prediction of ABX3 Perovskite Formation Energy Using Machine Learning,” Materials, vol. 18, no. 13, p. 2927, Jun. 2025, doi: 10.3390/ma18132927.
[15] R. M. Devadas and S. T, “Quantum machine learning: A comprehensive review of integrating AI with quantum computing for computational advancements,” MethodsX, vol. 14, p. 103318, Jun. 2025, doi: 10.1016/j.mex.2025.103318.
[16] M. Schuld and N. Killoran, “Quantum Machine Learning in Feature Hilbert Spaces,” Phys. Rev. Lett., vol. 122, no. 4, p. 040504, Feb. 2019, doi: 10.1103/PhysRevLett.122.040504.
[17] J. Wei et al., “Machine learning in materials science,” InfoMat, vol. 1, no. 3, pp. 338–358, Sep. 2019, doi: 10.1002/inf2.12028.
[18] S. Körbel, M. A. L. Marques, and S. Botti, “Stability and electronic properties of new inorganic perovskites from high-throughput ab initio calculations,” J. Mater. Chem. C Mater., vol. 4, no. 15, pp. 3157–3167, 2016, doi: 10.1039/C5TC04172D.
[19] C. Zuccarini, K. Ramachandran, and D. D. Jayaseelan, “Material discovery and modeling acceleration via machine learning,” APL Mater., vol. 12, no. 9, Sep. 2024, doi: 10.1063/5.0230677.
[20] A. Kesorn et al., “Formation energy prediction of neutral single-atom impurities in 2D materials using tree-based machine learning,” Mach. Learn. Sci. Technol., vol. 5, no. 3, p. 035039, Sep. 2024, doi: 10.1088/2632-2153/ad66ae.
[21] Z. Krunic, F. Flother, G. Seegan, N. Earnest-Noble, and S. Omar, “Quantum Kernels for Real-World Predictions Based on Electronic Health Records,” IEEE Transactions on Quantum Engineering, vol. 3, pp. 1–11, 2022, doi: 10.1109/TQE.2022.3176806.
[22] Z. Özpolat, Ö. Yıldırım, and M. Karabatak, “The Effect of Linear Discriminant Analysis and Quantum Feature Maps on QSVM Performance for Obesity Diagnosis,” Balkan Journal of Electrical and Computer Engineering, vol. 12, no. 3, pp. 206–213, Sep. 2024, doi: 10.17694/bajece.1475896.
[23] K.-L. Du, B. Jiang, J. Lu, J. Hua, and M. N. S. Swamy, “Exploring Kernel Machines and Support Vector Machines: Principles, Techniques, and Future Directions,” Mathematics, vol. 12, no. 24, p. 3935, Dec. 2024, doi: 10.3390/math12243935.
[24] V. S. Naresh and S. Reddi, “Quantum SVM-based predictive analytics: transforming classification methods in healthcare and beyond,” Quantum Inf. Process., vol. 24, no. 9, p. 272, Aug. 2025, doi: 10.1007/s11128-025-04892-6.
[25] A. Callison and N. Chancellor, “Hybrid quantum-classical algorithms in the noisy intermediate-scale quantum era and beyond,” Phys. Rev. A (Coll Park)., vol. 106, no. 1, p. 010101, Jul. 2022, doi: 10.1103/PhysRevA.106.010101.
[26] B. Sekeroglu, Y. K. Ever, K. Dimililer, and F. Al-Turjman, “Comparative Evaluation and Comprehensive Analysis of Machine Learning Models for Regression Problems,” Data Intell., vol. 4, no. 3, pp. 620–652, Jul. 2022, doi: 10.1162/dint_a_00155.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Fauzan Arif Prayogi, Muhamad Akrom
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License (Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) ) that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
