Neural Architecture Search and Reinforcement Learning for Adaptive Software Component Optimization in Engineering Workflows

Murasaki Shōnagon Natsume

Authors

Murasaki Shōnagon Natsume Machine Learning Systems Engineer, Japan Author

Keywords:

Neural Architecture Search, Reinforcement Learning, Software Optimization, Engineering Workflow, Adaptation, Deep Learning, Workflow Scheduling

Abstract

Engineering workflows are evolving rapidly, becoming increasingly data-driven, distributed, and modular. However, static software configurations often struggle to adapt to such dynamic execution environments. This research explores a novel integration of Neural Architecture Search (NAS) and Reinforcement Learning (RL) to enable dynamic optimization of software components within these workflows.

We propose a unified framework where NAS automates neural model design for workflow tasks, while RL governs adaptive decision-making in real time. Evaluated across multiple engineering workflow simulations, our hybrid approach demonstrated improvements in execution efficiency, fault tolerance, and adaptability. This work contributes to the growing movement toward intelligent, self-optimizing systems in software engineering.

References

Melnik, M., Nasonov, D. (2019). Workflow scheduling using neural networks and reinforcement learning. Procedia Computer Science, 150, 268–275.

Wistuba, M., Rawat, A., Pedapati, T. (2019). A survey on neural architecture search. arXiv preprint arXiv:1905.01392.

Gummadi, V. P. K. (2019). Microservices architecture with APIs: Design, implementation, and MuleSoft integration. Journal of Electrical Systems, 15(4), 130–134. https://doi.org/10.52783/jes.9328

Naik, K.J., Pedagandam, M. (2021). Workflow scheduling optimisation using neural networks and reinforcement learning. International Journal of Computer Science and Engineering, 23(2), 105–119.

Balaprakash, P., Egele, R., Salim, M., Wild, S. (2019). Scalable reinforcement-learning-based NAS for cancer research. Proceedings of the ACM/IEEE SC Conference, 1–12.

Nikitin, N.O., Vychuzhanin, P., Sarafanov, M. (2022). Automated evolutionary design of composite machine learning pipelines. Future Generation Computer Systems, 135, 40–56.

Esfahani, N., Elkhodary, A., Malek, S. (2013). Learning-based framework for self-adaptive software systems. IEEE Transactions on Software Engineering, 39(11), 1467–1493.

Gummadi, V. P. K. (2020). API design and implementation: RAML and OpenAPI specification. Journal of Electrical Systems, 16(4). https://doi.org/10.52783/jes.9329

Serban, A., Visser, J. (2022). Adapting software architectures to ML challenges. IEEE International Conference on Software Architecture, 1–10.

Wu, N., Xie, Y. (2022). Machine learning for computer architecture and systems. ACM Computing Surveys, 54(7), 1–36.

Richardson, N., Kothapalli, S., Onteddu, A.R. (2023). AI-driven optimization for evolving software architectures. ABC Journal of AI Engineering, 8(4), 301–319.

Anasuri, S., Rusum, G.P., Pappula, K.K. (2023). AI-driven software design patterns: Automation in system architecture. International Journal of Advances in Data Science and Machine Learning, 5(1), 80–95.