Pengaruh Feature Engineering Time Series Berbasis Rolling Statistics untuk Prediksi Suhu Menggunakan Random Forest dan Gradient Boosting
DOI:
https://doi.org/10.32493/jiup.v11i1.58137Keywords:
Rolling Statistics Feature Engineering, Random Forest Regression Model, Air Temperature Prediction, Time Series Temperature Prediction, Walk-Forward Validation MethodAbstract
Temperature data are a form of time series characterized by autocorrelation, seasonality, and nonlinear fluctuations, thus requiring appropriate temporal feature representation to achieve accurate predictions. This study aims to analyze the effect of rolling statistics–based feature engineering on temperature prediction performance using the multivariate Jena Climate dataset. Temporal features are constructed through a combination of lag features and rolling statistics (rolling mean and rolling standard deviation) with various window sizes. Prediction models are developed using Random Forest as the primary model and Gradient Boosting as a comparative model, evaluated using a walk-forward validation scheme and assessed using MAE and RMSE. The results show that the application of rolling statistics produces relatively stable performance but does not lead to a significant improvement in prediction accuracy. In the Random Forest model, the MAE increases from 0.0214 to 0.0216 and the RMSE from 0.1277 to 0.1295 after incorporating rolling features. These findings indicate that rolling statistics play a greater role in maintaining the stability of feature representation, while their contribution to improving predictive accuracy remains limited.
References
H. Zhang, Y. Liu, C. Zhang, and N. Li, “Machine Learning Methods for Weather Forecasting: A Survey,” Atmosphere (Basel)., vol. 16, no. 1, pp. 1–34, 2025, doi: https://doi.org/10.3390/atmos16010082.
H. M. Alkhezi, “A comparative analysis of forecasting models for air temperature: Classical time series, machine learning, and functional approaches,” Kuwait J. Sci., vol. 53, no. 1, p. 100529, 2026, doi: https://doi.org/10.1016/j.kjs.2025.100529.
M. Fauzi et al., “Implementasi Machine Learning Untuk Memprediksi Cuaca Menggunakan Support Vector Machine,” J. Ilm. Komputasi, vol. 23, no. 1, pp. 45–50, 2024, doi: https://doi.org/10.32409/jikstik.23.1.3499.
J. Kim, H. Kim, H. G. Kim, D. Lee, and S. Yoon, “A comprehensive survey of deep learning for time series forecasting: architectural diversity and open challenges,” Artif. Intell. Rev., vol. 58, no. 7, 2025, doi: https://doi.org/10.1007/s10462-025-11223-9.
C. Tan, J. Zhong, D. Yang, and W. Huang, “Forecasting and analysing global average temperature trends based on LSTM and ARIMA models,” PLoS One, vol. 20, no. 9 September, pp. 1–27, 2025, doi: https://doi.org/10.1371/journal.pone.0330645.
Y. Wang, H. Wu, J. Dong, Y. Liu, M. Long, and J. Wang, “Deep Time Series Models: A Comprehensive Survey and Benchmark,” arXiv preprint arXiv:2407.13278, 2024, doi: 10.48550/arXiv.2407.13278.
N. P. Nur Fauzi, S. Khomsah, and A. D. Putra Wicaksono, “Penerapan Feature Engineering dan Hyperparameter Tuning untuk Meningkatkan Akurasi Model Random Forest pada Klasifikasi Risiko Kredit,” J. Teknol. Inf. dan Ilmu Komput., vol. 12, no. 2, pp. 251–262, 2025, doi: https://doi.org/10.25126/jtiik.2025128472.
X. Ma and H. Zhang, “Time Series Forecasting Method Based on Multi-Scale Feature Fusion and Autoformer,” Appl. Sci., vol. 15, no. 7, 2025, doi: https://doi.org/10.3390/app15073768.
M. Shafiuzzaman et al., “Enhanced very short-term load forecasting with multi-lag feature engineering and prophet-XGBoost-CatBoost architecture,” Energy, vol. 335, no. February 2024, p. 137981, 2025, doi: https://doi.org/10.1016/j.energy.2025.137981.
B. Khaw, R. Irwanto, R. Yunis, and E. Elly, “Analisis Time Series dan Perancangan Dashboard untuk Memprediksi Penjualan dengan Metode Prophet dan SARIMAX,” J. Sifo Mikroskil, vol. 26, no. 2, 2025, doi: https://doi.org/10.55601/jsm.v26i2.1797.
N. A. Elmunim et al., “Enhancing Wind and Solar Energy Forecasting Through Time-Series Feature Engineering and Ensemble Machine Learning,” Scientific Reports, vol. 16, 2026, doi: https://doi.org/10.1038/s41598-026-49373-7.
S. A. F. Mortezanejad and R. Wang, “Addressing Challenges in Time Series Forecasting: A Comprehensive Comparison of Machine Learning Techniques,” no. March 2025, 2025, [Online]. Available: http://arxiv.org/abs/2503.20148
P. Arsi, T. Astuti, D. Rahmawati, and P. Subarkah, “Implementasi Sliding Window Algotihm pada Prediksi Kurs berbasis Neural Network,” DoubleClick J. Comput. Inf. Technol., vol. 6, no. 1, p. 51, 2022, doi: https://doi.org/10.25273/doubleclick.v6i1.13496.
E. P. Wahyuddin, R. E. Caraka, R. Kurniawan, W. Caesarendra, P. U. Gio, and B. Pardamean, “Improved LSTM hyperparameters alongside sentiment walk-forward validation for time series prediction,” J. Open Innov. Technol. Mark. Complex., vol. 11, no. 1, p. 100458, 2025, doi: https://doi.org/10.1016/j.joitmc.2024.100458.
C. Wang, M. D. Arseniev-Koehler, and M. Schmidt, “Towards Time-Series Feature Engineering in Automated Machine Learning for Multi-Step-Ahead Forecasting,” Engineering Proceedings, vol. 18, no. 1, p. 17, 2022, doi: https://doi.org/10.3390/engproc2022018017.
S. A. Khoiri and A. Wahid, “Jurnal Sistem dan Teknologi Informasi Indonesia Analisis Kinerja Algoritma Machine Learning dalam Prediksi Harga Cryptocurrency Performance Analysis of Machine Learning Algorithms in Cryptocurrency Price Prediction,” vol. 9, no. 2, pp. 133–141, 2024, doi: https://doi.org/10.32528/justindo.v9i2.1965.
A. Sutaryani, S. Sunarno, dan D. Djuniadi, “Perbandingan Performa Model Machine Learning dalam Prediksi Suhu di Semarang,” Jurnal Informatika dan Teknik Elektro Terapan, vol. 12, no. 3, pp. 2770–2775, 2024, doi: https://doi.org/10.23960/jitet.v12i3.4884.
H. Leidiyana and S. Nurajizah, “Pendekatan Hybrid Statistik dan Machine Learning untuk Peramalan Jumlah Kunjungan Turis,” J. Komtika (Komputasi dan Inform., vol. 9, no. 2, pp. 184–194, 2025, doi: https://doi.org/10.31603/komtika.v9i2.14909.
S. Karmaker, “Feature Engineering in Time Series Forecasting,” no. November, 2025, doi: https://doi.org/10.14293/PR2199.002316.v1.
L. Alpianto, A. Hermawan, and Junaedi, “Moving Average untuk Prediksi Harga Saham dengan Linear Regression,” J. Buana Inform., vol. 14, no. 02, pp. 117–126, 2023, doi: https://doi.org/10.24002/jbi.v14i02.7446.
Downloads
Published
License
Copyright (c) 2026 Czidni Sika Azkia, Yulian Ansori, Febriyanti Darnis, Arief Rahman
This work is licensed under a Creative Commons Attribution-NonCommercial 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-NonCommercial 4.0 International (CC BY-NC 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).
Jurnal Informatika Universitas Pamulang have CC-BY-NC or an equivalent license as the optimal license for the publication, distribution, use, and reuse of scholarly work.
In developing strategy and setting priorities, Jurnal Informatika Universitas Pamulang recognize that free access is better than priced access, libre access is better than free access, and libre under CC-BY-NC or the equivalent is better than libre under more restrictive open licenses. We should achieve what we can when we can. We should not delay achieving free in order to achieve libre, and we should not stop with free when we can achieve libre.
Jurnal Informatika Universitas Pamulang is licensed under a Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
YOU ARE FREE TO:
- Share : copy and redistribute the material in any medium or format
- Adapt : remix, transform, and build upon the material for any purpose, even commercially.
- The licensor cannot revoke these freedoms as long as you follow the license terms
This work is licensed under a 