Graph Convolutional Network (GCN) Untuk Klasifikasi Konektivitas Bandara

Rendy Saputro, Dian Puspita Hapsari

Abstract

Jaringan penerbangan global merupakan sistem transportasi kompleks yang secara alamiah membentuk struktur graf keruangan (spasial). Pemahaman terhadap hierarki konektivitas bandara sangat krusial bagi analisis rute dan optimalisasi infrastruktur. Namun, pemodelan data relasional berstruktur graf memiliki tantangan tersendiri jika diselesaikan menggunakan pendekatan machine learning konvensional berskala tabular. Penelitian ini mengusulkan penerapan arsitektur Deep Learning berbasis graf, yaitu Graph Convolutional Network (GCN), untuk mengklasifikasikan tingkat konektivitas bandara di seluruh dunia. Dataset yang digunakan merupakan rekaman operasional penerbangan global tahun 2024 yang .merepresentasikan 7.698 simpul (bandara) dan 36.589 sisi (edge) unik antarbandara, yang dibentuk dari 66.316 baris data rute mentah (satu baris per maskapai per rute) setelah melalui tahap pembersihan data.. Karena dataset tidak memiliki anotasi kelas yang eksplisit, algoritma Fuzzy C-Means (FCM) diterapkan secara unsupervised berdasarkan metrik degree centrality untuk menghasilkan pseudo-ground truth. Melalui metode ini, bandara diklasifikasikan ke dalam tiga kategori: Rendah, Sedang, dan Tinggi. Arsitektur GCN dirancang menggunakan dua lapisan konvolusi graf dengan fungsi aktivasi ReLU, teknik regularisasi Dropout dengan rasio 0.5, serta fungsi keluaran klasifikasi Softmax. Pembagian data dilakukan menggunakan skema transductive node masking dengan rasio 70% data latih, 15% data validasi, dan 15% data uji guna menjaga keutuhan topologi jaringan. Hasil pengujian menunjukkan bahwa model GCN yang dioptimasi menggunakan algoritma Adam pada 900 Epoch mampu memprediksi kelas konektivitas dengan tingkat akurasi keseluruhan sebesar 95% dan Macro F1-Score sebesar 79%. Secara keseluruhan, penelitian ini membuktikan bahwa algoritma GCN efektif dan tangguh dalam menangkap dan mengagregasi fitur atribut spasial secara simultan dengan topologi koneksi untuk pemodelan node classification pada sistem transportasi udara skala besar yang tidak seimbang (imbalanced data).

Keywords

Graph Convolutional Network, Konektivitas Bandara, Fuzzy C-Means, Klasifikasi Node, Jaringan Penerbangan.

Full Text:

PDF

References

Bandara, D., and Riccardi, K. 2024. Graph Node Classification to Predict Autism Risk in Genes. Genes, 15(4): 447-458.

Bezdek, J.C. 1981. Pattern Recognition with Fuzzy Objective Function Algorithms. New York: Plenum Press.

Chen, H., Huang, Z., Xu, Y., Deng, Z., Huang, F., He, P., and Li, Z. 2022. Neighbor Enhanced Graph Convolutional Networks for Node Classification and Recommendation. arXiv preprint, arXiv:2203.16097.

Dunn, J.C. 1973. A Fuzzy Relative of the ISODATA Process and Its Use in Detecting Compact Well-Separated Clusters. Journal of Cybernetics, 3(3): 32-57.

Goodfellow, I., Bengio, Y., and Courville, A. 2016. Deep Learning. Cambridge: MIT Press.

Guimerà, R., Mossa, S., Turtschi, A., and Amaral, L.A.N. 2005. The World-Wide Air Transportation Network: Anomalous Centrality, Community Structure, and Cities' Global Roles. Proceedings of the National Academy of Sciences, 102(22): 7794-7799.

Hammond, D.K., Vandergheynst, P., and Gribonval, R. 2011. Wavelets on Graphs via Spectral Graph Theory. Applied and Computational Harmonic Analysis, 30(2): 129-150.

Huang, Y., Yang, H., and Yan, Z. 2025. A Dynamic Multi-Graph Convolutional Spatial-Temporal Network for Airport Arrival Flow Prediction. Aerospace, 12(5): 395-410.

Jiang, W., and Luo, J. 2022. Graph Neural Network for Traffic Forecasting: A Survey. Expert Systems with Applications, 207: 117921.

Kadhim Abdulzahra, A.M., and Al-Qurabat, A.K.M. 2022. A Clustering Approach Based on Fuzzy C-Means in Wireless Sensor Networks for IoT Applications. Karbala International Journal of Modern Science, 8(4): 579-595.

Kingma, D.P., and Ba, J. 2017. Adam: A Method for Stochastic Optimization. arXiv preprint, arXiv:1412.6980.

Kipf, T.N., and Welling, M. 2017. Semi-Supervised Classification with Graph Convolutional Networks. Proceedings of the International Conference on Learning Representations (ICLR), hal. 1-14.

Nair, V., and Hinton, G.E. 2010. Rectified Linear Units Improve Restricted Boltzmann Machines. Proceedings of the 27th International Conference on Machine Learning (ICML), hal. 807-814.

Rafiee, A. 2024. Airports, Airlines, Planes, and Routes Update 2024 Dataset. Kaggle Repository.

Russell, S., and Norvig, P. 2020. Artificial Intelligence: A Modern Approach Fourth Edition. Hoboken: Pearson Education.

Seddik, M.E.A., Wu, C., Lutzeyer, J.F., and Vazirgiannis, M. 2022. Node Feature Kernels Increase Graph Convolutional Network Robustness. arXiv preprint, arXiv:2109.01785.

Sejan, M.A.S., Rahman, M.H., Aziz, M.A., Baik, J.I., You, Y.H., and Song, H.K. 2023. Graph Convolutional Network Design for Node Classification Accuracy Improvement. Mathematics, 11(17): 3680-3695.

Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., and Salakhutdinov, R. 2014. Dropout: A Simple Way to Prevent Neural Networks from Overfitting. Journal of Machine Learning Research, 15(1): 1929-1958.

Sun, X., and Wandelt, S. 2021. Robustness of Air Transportation as Complex Networks: Systematic Review of 15 Years of Research and Outlook into the Future. Sustainability, 13(11): 6446-6468.

Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., and Yu, P.S. 2019. A Comprehensive Survey on Graph Neural Networks. IEEE Transactions on Neural Networks and Learning Systems, 32(1): 4-24.

Zang, H., Zhu, J., and Gao, Q. 2022. Deep Learning Architecture for Flight Flow Spatiotemporal Prediction in Airport Network. Electronics, 11(23): 4058-4071.

Zanin, M., and Lillo, F. 2013. Modelling the Air Transport with Complex Networks: A Short Review. The European Physical Journal Special Topics, 215(1): 5-21.

Zhou, J., Cui, G., Hu, S., Zhang, Z., Yang, C., Liu, Z., Wang, L., Li, C., and Sun, M. 2020. Graph Neural Networks: A Review of Methods and Applications. AI Open, 1: 57-81.

DOI: https://doi.org/10.31284/j.integer.2026.v11i2.9353

Refbacks

  • There are currently no refbacks.