Perbandingan Fungsi Aktivasi Untuk Meningkatkan Kinerja Model LSTM Dalam Prediksi Ketinggian Air Sungai

Kanada Kurniawan, Barry Ceasaro, Sucipto Sucipto

Abstract


Sungai Kapuas memegang peran esensial dalam dinamika ekonomi masyarakat di Kota Pontianak, memberikan kontribusi signifikan dalam sektor transportasi, pertanian, serta pengelolaan tata air dan mitigasi banjir. Meskipun begitu, Sungai Kapuas juga memberikan tantangan serius terkait banjir yang kerap terjadi. Wilayah sepanjang Sungai Kapuas secara resmi ditetapkan sebagai Wilayah Sungai Strategis Nasional, menandakan pentingnya peran ekosistem sungai ini dalam konteks strategis nasional. Penelitian ini mengeksplorasi penerapan metode Long Short-Term Memory (LSTM) dengan berbagai fungsi aktivasi untuk memprediksi ketinggian air Sungai Kapuas. Dua belas jenis fungsi aktivasi, termasuk Sigmoid , Tanh, Hard Sigmoid , ELU, Exponential, GELU, Mish, ReLU, SELU, Swish, Softplus, dan Softsign, dievaluasi menggunakan lima metrik kinerja, yaitu Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Nash–Sutcliffe Efficiency (NSE), Pearson Correlation Coefficient (PCC), dan Kling-Gupta Efficiency (KGE). Data yang digunakan merupakan data ketinggian air Sungai Kapuas hasil observasi tiap jam dari Stasiun Meteorologi Maritim Pontianak selama periode 2016 hingga 2022. Hasil evaluasi menunjukkan bahwa fungsi aktivasi ReLU memberikan kinerja terbaik dengan nilai r, NSE, KGE, MAE, dan RMSE secara berurutan sebesar 0.990594, 0.981103, 0.988685, 3.78747, dan 5.2546.


Keywords


LSTM; Activation Function; ReLU; Ketinggian Air; Sungai Kapuas

Full Text:

PDF

References


F. Göltenboth dan W. Erdelen, “Geography And Geology,” dalam Ecology of Insular Southeast Asia, Elsevier, 2006, hlm. 3–16. doi: 10.1016/B978-044452739-4/50002-4.

K. D. Yanti, “Water Quality Analysis And Pollution Load Capacity Of River Kapuas In Mukok District, Sanggau Regency,” JTLLB, vol. 10, no. 1, hlm. 022, Agu 2021, doi: 10.26418/jtllb.v10i1.48610.

W. Widjonarko, I. A. Ardi, dan S. R. Hidayati, “Bencana Multi Bahaya Pada Daerah Aliran Sungai Kapuas, Kalimantan Barat,” Reka Ruang, vol. 4, no. 2, 2021, doi: 10.33579/rkr.v4i2.2756.

Menteri Pekerjaan Umum, Pola Pengelolaan Sumber Daya Air Wilayah Sungai Kapuas.

A. Rahmayani, Y. Darmadi, dan A. Firmansyah, Dari Hulu ke Hilir: Integrasi Ekonomi di Sungai Kapuas Pada 1900-1942, Cetakan pertama. Baturetno, Banguntapan, Yogyakarta: Diva Press, 2018.

H. D. Nugraha, “Implementasi Penanggulangan Banjir Rob Melalui Program Penyelenggaraan Drainase Di Kecamatan Pontianak Selatan Oleh Dinas Pekerjaan Umum Dan Penataan Ruang Kota Pontianak Provinsi Kalimantan Barat,” IPDN Jatinangor, 2021. [Daring]. Tersedia pada: http://eprints.ipdn.ac.id/6199/

J. Sampurno, V. Vallaeys, R. Ardianto, dan E. Hanert, “Modeling interactions between tides, storm surges, and river discharges in the Kapuas River delta,” Biogeosciences, vol. 19, no. 10, hlm. 2741–2757, Jun 2022, doi: 10.5194/bg-19-2741-2022.

R. Ardianto, A. Ismanto, J. Sampurno, dan S. Widada, “Tidal Flood Model Projection Using Land Subsidence Parameter In Pontianak, Indonesia,” GT, vol. 17, no. 2/2022, hlm. 135–147, Okt 2022, doi: 10.21163/GT_2022.172.12.

D. Kim, H. Han, W. Wang, dan H. Soo Kim, “Improvement of Deep Learning Models for River Water Level Prediction Using Complex Network Method,” Water, vol. 14, no. 3, hlm. 469, 2022, doi: 10.3390/w14030469.

J. F. Ruma, M. S. G. Adnan, A. Dewan, dan R. M. Rahman, “Particle swarm optimization based LSTM networks for water level forecasting: A case study on Bangladesh river network,” Results in Engineering, vol. 17, hlm. 100951, Mar 2023, doi: 10.1016/j.rineng.2023.100951.

C.-H. Yang, C.-H. Wu, dan C.-M. Hsieh, “Long Short-Term Memory Recurrent Neural Network for Tidal Level Forecasting,” IEEE Access, vol. 8, hlm. 159389–159401, 2020, doi: 10.1109/ACCESS.2020.3017089.

I. Fakhruzi, A. C. Siregar, P. Y. Utami, dan R. F. Leona, “Perbandingan Model Arima dan Deep Learning untuk Peramalan Kasus Covid-19,” CYBERNETICS, vol. 6, no. 02, hlm. 97–102, 2022.

V. K. dan S. K., “Towards activation function search for long short-term model network: A differential evolution based approach,” Journal of King Saud University - Computer and Information Sciences, vol. 34, no. 6, hlm. 2637–2650, Jun 2022, doi: 10.1016/j.jksuci.2020.04.015.

S. K. Das, P. Ranjan, P. Khan, dan S. Kumar, “log-Sigmoid Activation-Based Long Short-Term Memory for Time Series Data Classification,” IEEE Trans. Artif. Intell., hlm. 1–12, 2023, doi: 10.1109/TAI.2023.3265641.

A. Farzad, H. Mashayekhi, dan H. Hassanpour, “A comparative performance analysis of different activation functions in LSTM networks for classification,” Neural Comput & Applic, vol. 31, no. 7, hlm. 2507–2521, Jul 2019, doi: 10.1007/s00521-017-3210-6.

T. Szandała, “Review and Comparison of Commonly Used Activation Functions for Deep Neural Networks,” dalam Bio-inspired Neurocomputing, vol. 903, A. K. Bhoi, P. K. Mallick, C.-M. Liu, dan V. E. Balas, Ed., Singapore: Springer Singapore, 2021, hlm. 203–224. doi: 10.1007/978-981-15-5495-7_11.

M. H. Essai Ali, A. B. Abdel-Raman, dan E. A. Badry, “Developing Novel Activation Functions Based Deep Learning LSTM for Classification,” IEEE Access, vol. 10, hlm. 97259–97275, 2022, doi: 10.1109/ACCESS.2022.3205774.

T. Chai dan R. R. Draxler, “Root mean square error (RMSE) or mean absolute error (MAE)? – Arguments against avoiding RMSE in the literature,” Geosci. Model Dev., vol. 7, no. 3, hlm. 1247–1250, Jun 2014, doi: 10.5194/gmd-7-1247-2014.

J. Li, Y. Wang, Y. Li, W. Ming, Y. Long, dan M. Zhang, “Relationship between meteorological and hydrological droughts in the upstream regions of the Lancang–Mekong River,” Journal of Water and Climate Change, vol. 13, no. 2, hlm. 421–433, Feb 2022, doi: 10.2166/wcc.2021.445.

M. P. Clark dkk., “The Abuse of Popular Performance Metrics in Hydrologic Modeling,” Water Resources Research, vol. 57, no. 9, hlm. e2020WR029001, Sep 2021, doi: 10.1029/2020WR029001.

H. V. Gupta, H. Kling, K. K. Yilmaz, dan G. F. Martinez, “Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling,” Journal of Hydrology, vol. 377, no. 1–2, hlm. 80–91, Okt 2009, doi: 10.1016/j.jhydrol.2009.08.003.

R. M. Schmidt, “Recurrent Neural Networks (RNNs): A gentle Introduction and Overview,” 2019, doi: 10.48550/ARXIV.1912.05911.

Y. Yu, X. Si, C. Hu, dan J. Zhang, “A Review of Recurrent Neural Networks: LSTM Cells and Network Architectures,” Neural Computation, vol. 31, no. 7, hlm. 1235–1270, Jul 2019, doi: 10.1162/neco_a_01199.

D. I. Puteri, “Implementasi Long Short Term Memory (LSTM) dan Bidirectional Long Short Term Memory (BiLSTM) Dalam Prediksi Harga Saham Syariah,” Euler J. Ilm. Mat. Sains dan Teknol., vol. 11, no. 1, hlm. 35–43, Mei 2023, doi: 10.34312/euler.v11i1.19791.

C. Nwankpa, W. Ijomah, A. Gachagan, dan S. Marshall, “Activation Functions: Comparison of trends in Practice and Research for Deep Learning,” 2018, doi: 10.48550/ARXIV.1811.03378.

W. Kirch, Ed., “Pearson’s Correlation Coefficient,” dalam Encyclopedia of Public Health, Dordrecht: Springer Netherlands, 2008, hlm. 1090–1091. doi: 10.1007/978-1-4020-5614-7_2569.

L. Duc dan Y. Sawada, “A signal-processing-based interpretation of the Nash–Sutcliffe efficiency,” Hydrol. Earth Syst. Sci., vol. 27, no. 9, hlm. 1827–1839, Mei 2023, doi: 10.5194/hess-27-1827-2023.

D. N. Moriasi, M. W. Gitau, N. Pai, dan P. Daggupati, “Hydrologic and Water Quality Models: Performance Measures and Evaluation Criteria,” Trans. ASABE, vol. 58, no. 6, hlm. 1763–1785, Des 2015, doi: 10.13031/trans.58.10715.

W. J. M. Knoben, J. E. Freer, dan R. A. Woods, “Technical note: Inherent benchmark or not? Comparing Nash–Sutcliffe and Kling–Gupta efficiency scores,” Hydrol. Earth Syst. Sci., vol. 23, no. 10, hlm. 4323–4331, Okt 2019, doi: 10.5194/hess-23-4323-2019.

L. Santos, G. Thirel, dan C. Perrin, “Technical note: Pitfalls in using log-transformed flows within the KGE criterion,” Hydrol. Earth Syst. Sci., vol. 22, no. 8, hlm. 4583–4591, Agu 2018, doi: 10.5194/hess-22-4583-2018.

B. Babcock, S. Babu, M. Datar, R. Motwani, dan J. Widom, “Models and issues in data stream systems,” dalam Proceedings of the twenty-first ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems, Madison Wisconsin: ACM, Jun 2002, hlm. 1–16. doi: 10.1145/543613.543615.

V. Braverman, “Sliding Window Algorithms,” dalam Encyclopedia of Algorithms, M.-Y. Kao, Ed., New York, NY: Springer New York, 2016, hlm. 2006–2011. doi: 10.1007/978-1-4939-2864-4_797.

K. Kurniawan, J. Sampurno, R. Adriat, R. Ardianto, dan A. A. Kushadiwijayanto, “Deep-Learning-Based LSTM Model for Predicting a Tidal River’s Water Levels: A Case Study of the Kapuas Kecil River, Indonesia,” dalam Data Science and Artificial Intelligence, vol. 1942, C. Anutariya dan M. M. Bonsangue, Ed., Singapore: Springer Nature Singapore, 2023, hlm. 103–110. doi: 10.1007/978-981-99-7969-1_8.

P. Ramachandran, B. Zoph, dan Q. V. Le, “Searching for Activation Functions,” 2017, doi: 10.48550/ARXIV.1710.05941.

Y. Liu, J. Zhang, C. Gao, J. Qu, dan L. Ji, “Natural-Logarithm-Rectified Activation Function in Convolutional Neural Networks,” 2019, doi: 10.48550/ARXIV.1908.03682.




DOI: http://dx.doi.org/10.26418/jp.v10i1.72866

Refbacks

  • There are currently no refbacks.