Impact of DTW-Based K-Medoids and Fuzzy C-Means Clustering on the Forecasting Accuracy of ARIMA, TCN, and Hybrid Models in Anomalous Time Series

Meavi Cintani; I Made Sumertajaya; Yenni Angraini

doi:10.59395/ijadis.v7i2.1569

Authors

Meavi Cintani Statistics and Science Data, School of Data Science Mathematics and Informatics, IPB University, Indonesia https://orcid.org/0009-0006-9993-1746
I Made Sumertajaya Statistics and Science Data, School of Data Science Mathematics and Informatics, IPB University, Indonesia https://orcid.org/0000-0002-2682-7941
Yenni Angraini Statistics and Science Data, School of Data Science Mathematics and Informatics, IPB University, Indonesia https://orcid.org/0000-0003-3186-2378

DOI:

https://doi.org/10.59395/ijadis.v7i2.1569

Keywords:

Time series clustering, DTW-MDS, K-Medoid, Fuzzy C Means , Hybrid ARIMA-TCN, Export Forecasting

Abstract

This study investigates the impact of Dynamic Time Warping (DTW)-based clustering on forecasting accuracy in anomalous time series data. Monthly export values from 30 provinces in Indonesia are clustered using K-Medoids and Fuzzy C-Means (FCM) based on DTW distances reduced through Multidimensional Scaling (MDS). Cluster validation results indicate that FCM demonstrates better stability under anomalous conditions, with a Silhouette Score of 0.62 and a Davies–Bouldin Index of 0.57, compared to K-Medoids with a Silhouette Score of 0.61 and a Davies–Bouldin Index of 0.58. Forecasting performance is systematically evaluated through an expanding-window scheme using standalone ARIMA and Temporal Convolutional Network (TCN) models as baselines against a hybrid ARIMA–TCN approach. The results show that the hybrid ARIMA–TCN model achieves the lowest Mean Absolute Percentage Error (MAPE) in clusters with relatively stable patterns by effectively combining linear and non-linear components. However, for clusters characterized by higher volatility and irregular patterns, the standalone TCN model yields better forecasting accuracy. Furthermore, FCM clustering produces better overall forecasting accuracy, with an average MAPE of 7.11%, compared to 8.42% for K-Medoids. The relatively small difference in evaluation results between clean and empirical data suggests that the proposed DTW–MDS–clustering–forecasting framework maintains consistent performance in the presence of anomalies. The final model is then applied to generate export forecasts for the next 12 periods.

414 145

Downloads

Download data is not yet available.

References

[1] B. P. Statistik, “Statistik Ekspor Impor Indonesia 2023,” 2023.

[2] Adela Nursafira, Annisa Syafriani, Sabrina Natasya Matondang, Dayu Rizki Nurhamda, and Dwita Sakuntala, “Analysis of Factors Affecting Export and Import Performance,” Indonesian Journal of Entrepreneurship and Startups, vol. 3, no. 1, pp. 21–32, Jan. 2025, doi: 10.55927/ijes.v3i1.13355. DOI: https://doi.org/10.55927/ijes.v3i1.13355

[3] T. Warren Liao, “Clustering of time series data—a survey,” Pattern Recognit., vol. 38, no. 11, pp. 1857–1874, Nov. 2005, doi: 10.1016/j.patcog.2005.01.025. DOI: https://doi.org/10.1016/j.patcog.2005.01.025

[4] T. Fu, “A review on time series data mining,” Eng. Appl. Artif. Intell., vol. 24, no. 1, pp. 164–181, Feb. 2011, doi: 10.1016/j.engappai.2010.09.007. DOI: https://doi.org/10.1016/j.engappai.2010.09.007

[5] M. Zhang, J. Guo, X. Li, and R. Jin, “Data-Driven Anomaly Detection Approach for Time-Series Streaming Data,” Sensors, vol. 20, no. 19, p. 5646, Oct. 2020, doi: 10.3390/s20195646. DOI: https://doi.org/10.3390/s20195646

[6] E. Keogh and C. A. Ratanamahatana, “Exact indexing of dynamic time warping,” Knowl. Inf. Syst., vol. 7, no. 3, pp. 358–386, Mar. 2005, doi: 10.1007/s10115-004-0154-9. DOI: https://doi.org/10.1007/s10115-004-0154-9

[7] A. N. Putri, N. Satyahadewi, and S. Aprizkiyandari, “Pengelompokan Provinsi di Indonesia Menggunakan Time Series Clustering pada Sektor Ekspor Nonmigas,” Jambura Journal of Mathematics, vol. 6, no. 1, pp. 16–22, Feb. 2024, doi: 10.37905/jjom.v6i1.21921. DOI: https://doi.org/10.37905/jjom.v6i1.21921

[8] Y. Chen et al., “Delineating urban functional areas with building-level social media data: A dynamic time warping (DTW) distance based k -medoids method,” Landsc. Urban Plan., vol. 160, pp. 48–60, Apr. 2017, doi: 10.1016/j.landurbplan.2016.12.001. DOI: https://doi.org/10.1016/j.landurbplan.2016.12.001

[9] H. Izakian, W. Pedrycz, and I. Jamal, “Fuzzy clustering of time series data using dynamic time warping distance,” Eng. Appl. Artif. Intell., vol. 39, pp. 235–244, Mar. 2015, doi: 10.1016/j.engappai.2014.12.015. DOI: https://doi.org/10.1016/j.engappai.2014.12.015

[10] P. J. F. Borg, I.; Groenen, Modern Multidimensional Scaling: Theory and Applications. Springer, 2005. doi: 10.1007/0-387-28981-X.

[11] G. Hyndman, R. J. and Athanasopoulos, Forecasting: Principles and Practice, 3rd ed. OTexts, 2021.

[12] G. M. Box, G. E. P. and Jenkins, Time Series Analysis: Forecasting and Control. San Francisco, CA: Holden-Day, 1970.

[13] C. W. Borovykh, A., Bohte, S., and Oosterlee, “Conditional time series forecasting with convolutional neural networks,” arXiv preprint, vol. arXiv:1703, 2017.

[14] T. Chakraborty, S. Chattopadhyay, and I. Ghosh, “Forecasting dengue epidemics using a hybrid methodology,” Physica A: Statistical Mechanics and its Applications, vol. 527, p. 121266, Aug. 2019, doi: 10.1016/j.physa.2019.121266. DOI: https://doi.org/10.1016/j.physa.2019.121266

[15] I. M. W. A. S. K. R. Embay. Sumertajaya, “Vector Autoregressive-Moving Average Imputation Algorithm for Handling Missing Data in Multivariate Time Series,” IAENG Int. J. Comput. Sci., vol. 5, no. 2, pp. 727–735, May 2023.

[16] D. Tsabitah, Y. Angraini, and I. M. Sumertajaya, “Implementation of Clustering Time Series with DTW to Clustering and Forecasting Rice Prices Each Provinces in Indonesia,” Inferensi, vol. 8, no. 1, p. 13, Mar. 2025, doi: 10.12962/j27213862.v8i1.21952. DOI: https://doi.org/10.12962/j27213862.v8i1.21952

[17] G. G. Ghiffary, E. D. D. Yanuari, K. A. Notodiputro, Y. Angraini, and L. N. A. Mualifah, “COMPARATIVE PERFORMANCE OF SARIMAX AND LSTM MODEL IN PREDICTING IMPORT QUANTITIES OF MILK, BUTTER, AND EGGS,” BAREKENG: Jurnal Ilmu Matematika dan Terapan, vol. 19, no. 1, pp. 407–418, Jan. 2025, doi: 10.30598/barekengvol19iss1pp407-418. DOI: https://doi.org/10.30598/barekengvol19iss1pp407-418

[18] A. Ostermann, A. Bajrami, and A. Bogensperger, “Short-term forecasting of German generation-based CO2 emission factors using parametric and non-parametric time series models,” Energy Informatics, vol. 7, no. 1, p. 2, Jan. 2024, doi: 10.1186/s42162-024-00303-9. DOI: https://doi.org/10.1186/s42162-024-00303-9

[19] J. (1994) Berndt, D. J., & Clifford, “Using dynamic time warping to find patterns in time series. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining,” in Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (AAAIWS’94), 1994, pp. 359--370.

[20] I. Borg and P. J. F. Groenen, Modern Multidimensional Scaling. New York, NY: Springer New York, 2005. doi: 10.1007/0-387-28981-X. DOI: https://doi.org/10.1007/0-387-28981-X

[21] S. Bai, J. Z. Kolter, and V. Koltun, “An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling,” 2018, [Online]. Available: http://arxiv.org/abs/1803.01271

[22] Peter J. Huber; Elvezio Ronchetti, Robust Statistics, 2nd edition. Hoboken, New Jersey: John Wiley & Sons, 2009. DOI: https://doi.org/10.1002/9780470434697

[23] P. D’Urso, L. De Giovanni, and R. Massari, “Robust fuzzy clustering of multivariate time trajectories,” International Journal of Approximate Reasoning, vol. 99, pp. 12–38, Aug. 2018, doi: 10.1016/j.ijar.2018.05.002. DOI: https://doi.org/10.1016/j.ijar.2018.05.002

[24] G. P. Zhang, “Time series forecasting using a hybrid ARIMA and neural network model,” Neurocomputing, vol. 50, pp. 159–175, Jan. 2003, doi: 10.1016/S0925-2312(01)00702-0. DOI: https://doi.org/10.1016/S0925-2312(01)00702-0

[25] D. U. Tsabitah, Y. Angraini, and I. M. Sumertajaya, “Spatiotemporal Clustering of Key Food Commodity Prices Using Multivariate Time Series,” International Journal of Advances in Data and Information Systems, vol. 6, no. 3, pp. 638–652, Dec. 2025, doi: 10.59395/ijadis.v6i3.1422. DOI: https://doi.org/10.59395/ijadis.v6i3.1422

[26] E. A. Maharaj, P. D’Urso, and J. Caiado, Time Series Clustering and Classification. Chapman and Hall/CRC, 2019. doi: 10.1201/9780429058264. DOI: https://doi.org/10.1201/9780429058264

[27] F. Petitjean, A. Ketterlin, and P. Gançarski, “A global averaging method for dynamic time warping, with applications to clustering,” Pattern Recognit., vol. 44, no. 3, pp. 678–693, Mar. 2011, doi: 10.1016/j.patcog.2010.09.013. DOI: https://doi.org/10.1016/j.patcog.2010.09.013