ML-Empowered Microservice Workload Prediction by Dual-Regularized Matrix Factorization DOI Creative Commons
Xiaoxuan Luo, Hong Shen, Wei Ke

и другие.

Applied Sciences, Год журнала: 2025, Номер 15(11), С. 5946 - 5946

Опубликована: Май 25, 2025

A technical challenge for workload prediction in microservice systems is how to capture both the dynamic features of and evolving dependencies among microservices. The existing work focused mainly on modeling without taking adequate account due their unpredictable temporal dynamics. To fill this gap, as an illustration bridging theory real-work solutions by integrating machine learning with data analysis, we propose a novel framework Temporality-Dependence Dual-Regularized Matrix Factorization (TDDRMF) combining matrix factorization regularization temporality dependencies. It models product dependency W feature X applying factorization, computes low-rank norm convex relaxation rank minimization. further enhance its adaptability variations real-time environments, deploy error detection update mechanism. Experiments Alibaba dataset show that TDDRMF achieves 18.5% lower RMSE than TAMF 10-step prediction, improving methods accuracy. In comparison ML-based methods, uses only 5% training data, it requires small fraction time.

Язык: Английский

Research on Image Recognition and Classification Algorithms in Cloud Computing Environment Based on Deep Neural Networks DOI Creative Commons

Zihang Jia

IEEE Access, Год журнала: 2025, Номер 13, С. 19728 - 19754

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

1
Estimating Application Performance in Container-Based Environments: A Cross-Domain Monitoring Approach DOI
Marcelo de Miranda Reis, Rodrigo Moreira, Larissa Ferreira Rodrigues Moreira

и другие.

Lecture notes on data engineering and communications technologies, Год журнала: 2025, Номер unknown, С. 89 - 100

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Energy-deadline optimization with minimal task failure aware task partitioning model in heterogeneous cloud computing framework DOI

K. N. Divyaprabha,

T. S. B. Sudarshan

Computers & Electrical Engineering, Год журнала: 2025, Номер 125, С. 110438 - 110438

Опубликована: Май 17, 2025

Язык: Английский

Процитировано

0
ML-Empowered Microservice Workload Prediction by Dual-Regularized Matrix Factorization DOI Creative Commons
Xiaoxuan Luo, Hong Shen, Wei Ke

и другие.

Applied Sciences, Год журнала: 2025, Номер 15(11), С. 5946 - 5946

Опубликована: Май 25, 2025

A technical challenge for workload prediction in microservice systems is how to capture both the dynamic features of and evolving dependencies among microservices. The existing work focused mainly on modeling without taking adequate account due their unpredictable temporal dynamics. To fill this gap, as an illustration bridging theory real-work solutions by integrating machine learning with data analysis, we propose a novel framework Temporality-Dependence Dual-Regularized Matrix Factorization (TDDRMF) combining matrix factorization regularization temporality dependencies. It models product dependency W feature X applying factorization, computes low-rank norm convex relaxation rank minimization. further enhance its adaptability variations real-time environments, deploy error detection update mechanism. Experiments Alibaba dataset show that TDDRMF achieves 18.5% lower RMSE than TAMF 10-step prediction, improving methods accuracy. In comparison ML-based methods, uses only 5% training data, it requires small fraction time.

Язык: Английский

Процитировано

0