Implementing MLOps on Edge-Cloud Systems: A New Paradigm for Training at the Edge
Bibliographic record
Abstract
Owing to the rise in data from the Internet of Things~(IoT) devices and the increasing demand for intelligent decision-making on the network's edge, there has been a significant surge in interest in the intersection of edge computing, cloud computing and artificial intelligence~(AI). Various sectors are adopting such an integrated approach because of the low-latency operating capability due to edge computing, intelligent decision-making due to AI and scalable computing in the cloud. Due to low-latency requirements, in case of performance degradation of the AI application, it is crucial to rapidly adapt and update the edge environment independently while maintaining state synchronization with the cloud. \n \nOwing to the prerequisite for rapid adaptability, a necessity for personalized Machine Learning~(ML) training on the edge becomes evident. Furthermore, the universal ML model training is typically conducted in the cloud, leveraging its higher computing resources and abundant data in the central storage. In such a hybrid environment with multiple model sources, it is essential to maintain consistency and a synchronized state of the system. Conventional Machine Learning Operations, also known as MLOps, manage the efficient deployment and monitoring of machine learning models in a single-tier environment. \n \nThe challenge of performing MLOps in an edge-cloud environment grows with the number of IoT devices, edge servers and machine learning models. Thus, streamlining the machine learning process, including model training, deployment, and performance monitoring, requires a scalable and robust hybrid approach. To solve the challenge of performing multi-tiered MLOps in a hybrid ecosystem, we propose a novel MLOps architecture to orchestrate the edge-cloud model training and synchronization. \n \nThis thesis assesses the proposed architecture using quality attributes, including maintainability, reliability, scalability, functional adaptability and robustness. Furthermore, the thesis tests the proposed architecture in a practical case study experiment, including multiple IoT devices, edge servers and centralized cloud infrastructure. This thesis presents an innovative solution for maintaining ML-enabled edge-cloud systems.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.008 |
| Open science | 0.004 | 0.005 |
| Research integrity | 0.001 | 0.004 |
| Insufficient payload (model declined to judge) | 0.005 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".