KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting

Ezgi Nur Alışan; Emir Mehmet Eryilmaz; Emre Sefer; İsmail Aktürk

doi:10.35377/saucis...1941245

KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting

Abstract

Accurate forecasting of cloud resource utilization is essential for proactive provisioning, autoscaling, and energy-aware scheduling in modern data centers. Temporal Kolmogorov-Arnold Networks (TKANs) provide a promising alternative to MLP-based models by combining Kolmogorov- Arnold representations with recurrent sequential modeling. However, prior TKAN formulation typically relies on LSTM-style recurrence, which increases parameter count and computational cost. This work introduces KAN–GRU, a lightweight TKAN-based architecture that substitutes the LSTM cell in the TKAN temporal component with a GRU, reducing parameters while preserving temporal modeling capacity. The evaluation benchmarks KAN–GRU for data-center resource-utilization forecasting under univariate and multivariate configurations and across single-step and multi-step prediction horizons. The results indicate that KAN-GRU improves computational efficiency while maintaining competitive predictive performance, making it suitable for scalable deployment in monitoring and control operations in cloud environments.

Keywords

Project Number

BIDEB 2232-B Project No: 121C127

Ethical Statement

This study does not require ethics committee approval, as it does not involve human participants, animal subjects, personal data, or any experimental procedure requiring ethical review. It is declared that scientific and ethical principles were followed during the preparation of this study, and all sources used have been appropriately cited in the bibliography.

References

A. Katal, S. Dahiya, and T. Choudhury, “Energy efficiency in cloud computing data centers: A survey on software technologies,” Cluster Computing, vol. 26, no. 3, pp. 1845–1875, 2023, doi: 10.1007/s10586-022-03713-0.
S. M. Sajal et al., “Kerveros: Efficient and scalable cloud admission control,” in Proc. 17th USENIX Symp. Oper. Syst. Des. Implement. (OSDI), Boston, MA, USA, July 2023, pp. 227–245.
H. Yuan, J. Bi, S. Li, J. Zhang, and M. Zhou, “An improved LSTM-based prediction approach for resources and workload in large-scale data centers,” IEEE Internet of Things Journal, vol. 11, no. 12, pp. 22816–22829, June 2024, doi: 10.1109/JIOT.2024.3383512.
G. Marques, C. R. Senna, S. Sargento, L. Carvalho, L. Pereira, and R. Matos, “Proactive resource management for cloud of services environments,” Future Generation Computer Systems, vol. 150, pp. 90–102, Jan. 2024, doi: 10.1016/j.future.2023.08.005.
D. Saxena, J. Kumar, A. K. Singh, and S. Schmid, “Performance analysis of machine learning centered workload prediction models for cloud,” IEEE Transactions on Parallel and Distributed Systems, vol. 34, no. 4, pp. 1313–1330, Apr. 2023, doi: 10.1109/TPDS.2023.3240567.
M. P. Yadav, N. Pal, and D. K. Yadav, “Workload prediction over cloud server using time series data,” in Proc. 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence), Noida, India, Jan. 2021, pp. 267–272, doi: 10.1109/Confluence51648.2021.9377032.
Y. Dang, Q. Lin, and P. Huang, “AIOps: Real-world challenges and research innovations,” in Proc. 2019 IEEE/ACM 41st International Conference on Software Engineering: Companion Proceedings (ICSE-Companion), Montreal, QC, Canada, May 2019, pp. 4–5, doi: 10.1109/ICSE-Companion.2019.00023.
S. Malik, M. Tahir, M. Sardaraz, and A. Alourani, “A resource utilization prediction model for cloud data centers using evolutionary algorithms and machine learning techniques,” Applied Sciences, vol. 12, no. 4, Art. no. 2160, 2022, doi: 10.3390/app12042160.

B. Lim and S. Zohren, “Time-series forecasting with deep learning: A survey,” Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 379, no. 2194, Art. no. 20200209, 2021, doi: 10.1098/rsta.2020.0209.
K. Benidis et al., “Deep learning for time series forecasting: Tutorial and literature survey,” ACM Comput. Surv., vol. 55, no. 6, Art. no. 121, pp. 1–36, Dec. 2022, doi: 10.1145/3533382.
A. Rossi, A. Visentin, D. Carraro, S. Prestwich, and K. N. Brown, “Forecasting workload in cloud computing: Towards uncertainty-aware predictions and transfer learning,” Cluster Computing, vol. 28, Art. no. 258, 2025, doi: 10.1007/s10586-024-04933-2.
Z. Ahamed, M. Khemakhem, F. Eassa, F. Alsolami, and A. S. Al-Malaise Al-Ghamdi, “Technical study of deep learning in cloud computing for accurate workload prediction,” Electronics, vol. 12, no. 3, Art. no. 650, 2023, doi: 10.3390/electronics12030650.
J. Bi, H. Ma, H. Yuan, and J. Zhang, “Accurate prediction of workloads and resources with multi-head attention and hybrid LSTM for cloud data centers,” IEEE Trans. Sustain. Comput., vol. 8, no. 3, pp. 375–384, Jul./Sep. 2023, doi: 10.1109/TSUSC.2023.3259522.
C. He, M. Ma, and P. Wang, “Extract interpretability-accuracy balanced rules from artificial neural networks: A review,” Neurocomputing, vol. 387, pp. 346–358, 2020, doi: 10.1016/j.neucom.2020.01.036.
Z. Liu et al., “KAN: Kolmogorov–Arnold Networks,” 2024, arXiv:2404.19756.
S. Somvanshi, S. A. Javed, M. M. Islam, D. Pandit, and S. Das, “A survey on Kolmogorov–Arnold network,” ACM Comput. Surv., vol. 58, no. 2, Art. no. 55, pp. 1–35, 2026, doi: 10.1145/3743128.
C. Zeng, J. Wang, H. Shen, and Q. Wang, “KAN versus MLP on irregular or noisy function,” in Proc. 15th IEEE Int. Conf. Pattern Recognit. Syst. (ICPRS), Vi˜na del Mar, Chile, Dec. 2025, pp. 1–7, doi: 10.1109/ICPRS66293.2025.11302852.
R. Qiu, Y. Miao, S. Wang, Y. Zhu, L. Yu, and X.-S. Gao, “PowerMLP: An efficient version of KAN,” in Proc. AAAI Conf. Artif. Intell., vol. 39, no. 19, pp. 20069–20076, 2025, doi: 10.1609/aaai.v39i19.34210.
R. Genet and H. Inzirillo, “TKAN: Temporal Kolmogorov–Arnold Networks,” 2024, arXiv:2405.07344.
K. Cho et al., “Learning phrase representations using RNN encoder–decoder for statistical machine translation,” in Proc. 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), Doha, Qatar, Oct. 2014, pp. 1724–1734, doi: 10.3115/v1/D14-1179.
J. Chung, C. Gulcehre, K. Cho, and Y. Bengio, “Empirical evaluation of gated recurrent neural networks on sequence modeling,” 2014, arXiv:1412.3555.
R. N. Calheiros, E. Masoumi, R. Ranjan, and R. Buyya, “Workload prediction using ARIMA model and its impact on cloud applications’ QoS,” IEEE Transactions on Cloud Computing, vol. 3, no. 4, pp. 449–458, Oct.–Dec. 2015, doi: 10.1109/TCC.2014.2350475.
D. Janardhanan and E. Barrett, “CPU workload forecasting of machines in data centers using LSTM recurrent neural networks and ARIMA models,” in Proc. 12th International Conference for Internet Technology and Secured Transactions (ICITST), Cambridge, U.K., Dec. 2017, pp. 55–60, doi: 10.23919/ICITST.2017.8356346.
P. T. Yamak, L. Yujian, and P. K. Gadosey, “A comparison between ARIMA, LSTM, and GRU for time series forecasting,” in Proc. 2nd International Conference on Algorithms, Computing and Artificial Intelligence, Sanya, China, Dec. 2019, pp. 49–55, doi: 10.1145/3377713.3377722.
M. Duggan, K. Mason, J. Duggan, E. Howley, and E. Barrett, “Predicting host CPU utilization in cloud computing using recurrent neural networks,” in Proc. 12th International Conference for Internet Technology and Secured Transactions (ICITST), Cambridge, U.K., Dec. 2017, pp. 67–72, doi: 10.23919/ICITST.2017.8356348.
J. Dogani, F. Khunjush, M. R. Mahmoudi, and M. Seydali, “Multivariate workload and resource prediction in cloud computing using CNN and GRU by attention mechanism,” The Journal of Supercomputing, vol. 79, no. 3, pp. 3437–3470, 2023, doi: 10.1007/s11227-022-04782-z.
C. J. Vaca-Rubio, L. Blanco, R. Pereira, and M. Caus, “Kolmogorov–Arnold Networks (KANs) for time series analysis,” 2024, arXiv:2405.08790.
K. Xu, L. Chen, and S. Wang, “Kolmogorov–Arnold Networks for time series: Bridging predictive power and interpretability,” 2024, arXiv:2406.02496.
A. N. Kolmogorov, On the representation of continuous functions of several variables by superpositions of continuous functions of a smaller number of variables. American Mathematical Society, 1961.
R. Shi and C. Jiang, “Three-way ensemble prediction for workload in the data center,” IEEE Access, vol. 10, pp. 10021–10030, 2022, doi: 10.1109/AC- CESS.2022.3145426.
R. J. Hyndman and G. Athanasopoulos, Forecasting: Principles and Practice, 3rd ed. Melbourne, Australia: OTexts, 2021. [Online]. Available: https: //otexts.com/fpp3/
X. Qiu et al., “TFB: Towards comprehensive and fair benchmarking of time series forecasting methods,” Proc. VLDB Endow., vol. 17, no. 9, pp. 2363–2377, 2024, doi: 10.14778/3665844.3665863.
R. Genet and H. Inzirillo, “SigGate: Enhancing recurrent neural networks with signature-based gating mechanisms,” 2025, arXiv:2502.09318.
Alibaba Group, “Alibaba Cluster Trace v2018,” Alibaba Cluster Trace Program, Accessed: Aug. 2, 2026. [Online]. Available: https://github.com/alibaba/ clusterdata/tree/master/cluster-trace-v2018

Details

Primary Language

English

Subjects

Artificial Intelligence (Other)

Journal Section

Research Article

Authors

Ezgi Nur Alışan
0009-0003-9280-1074
Türkiye

Emir Mehmet Eryilmaz
0009-0006-1414-2474
Türkiye

Emre Sefer
0000-0002-9186-0270
Türkiye

İsmail Aktürk ^*
0000-0003-1970-2507
Türkiye

Early Pub Date

October 27, 2026

Publication Date

October 30, 2026

Submission Date

April 30, 2026

Acceptance Date

August 31, 2026

Published in Issue

Year 2026 Volume: 9 Number: 5

DOI

https://doi.org/10.35377/saucis...1941245

IZ

https://izlik.org/JA52ED58WJ

Cite

RIS / Bibtex

APA

Alışan, E. N., Eryilmaz, E. M., Sefer, E., & Aktürk, İ. (2026). KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting. Sakarya University Journal of Computer and Information Sciences, 9(5), 1571-1580. https://doi.org/10.35377/saucis...1941245

AMA

1.Alışan EN, Eryilmaz EM, Sefer E, Aktürk İ. KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting. SAUCIS. 2026;9(5):1571-1580. doi:10.35377/saucis.1941245

Chicago

Alışan, Ezgi Nur, Emir Mehmet Eryilmaz, Emre Sefer, and İsmail Aktürk. 2026. “KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting”. Sakarya University Journal of Computer and Information Sciences 9 (5): 1571-80. https://doi.org/10.35377/saucis. 1941245.

EndNote

Alışan EN, Eryilmaz EM, Sefer E, Aktürk İ (October 1, 2026) KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting. Sakarya University Journal of Computer and Information Sciences 9 5 1571–1580.

IEEE

[1]E. N. Alışan, E. M. Eryilmaz, E. Sefer, and İ. Aktürk, “KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting”, SAUCIS, vol. 9, no. 5, pp. 1571–1580, Oct. 2026, doi: 10.35377/saucis...1941245.

ISNAD

Alışan, Ezgi Nur - Eryilmaz, Emir Mehmet - Sefer, Emre - Aktürk, İsmail. “KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting”. Sakarya University Journal of Computer and Information Sciences 9/5 (October 1, 2026): 1571-1580. https://doi.org/10.35377/saucis. 1941245.

JAMA

1.Alışan EN, Eryilmaz EM, Sefer E, Aktürk İ. KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting. SAUCIS. 2026;9:1571–1580.

MLA

Alışan, Ezgi Nur, et al. “KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting”. Sakarya University Journal of Computer and Information Sciences, vol. 9, no. 5, Oct. 2026, pp. 1571-80, doi:10.35377/saucis. 1941245.

Vancouver

1.Ezgi Nur Alışan, Emir Mehmet Eryilmaz, Emre Sefer, İsmail Aktürk. KAN-GRU: A Lightweight Temporal Kolmogorov-Arnold Network for Cloud Resource Utilization Forecasting. SAUCIS. 2026 Oct. 1;9(5):1571-80. doi:10.35377/saucis. 1941245