Nowadays, there is a growing trend to deploy machine learning (ML) models on edge devices. To cope with the increasing resource requirements of current ML models, multi-accelerator edge devices that integrate CPU, GPU, NPU, or TPU in a single SoC gain popularity. However, we observe that existing ML inference serving frameworks are poor in utilizing the unique hardware architecture of these edge devices. In this paper, we present InFSCALER, an efficient ML inference serving framework tailored for multi-accelerator edge devices. InfSCALER discovers the architectural bottleneck of ML models and designs a bottleneck-aware asymmetric auto-scaling technique to facilitate efficient resource allocation for ML models on the edge. Furthermore, InfScaler capitalizes on the hardware’s unified memory feature inherent to edge devices, ensuring efficient data sharing between the asymmetrically scaled model partitions. Our experimental results show that InfScaler achieves up to $\mathbf{1 2 6. 5 9 \%}$ throughput improvement and $\mathbf{2 7. 3 2 \%}$ resource reduction while satisfying the latency requirements compared with the state-of-the-art inference serving approaches.

Federated graph classification has emerged as a promising paradigm for privacy-preserving graph learning across distributed clients. However, real-world federated scenarios often suffer from severe data heterogeneity and label noise, which significantly degrade model performance. To address these challenges, we propose FedRog, a robust and personalized federated graph neural network framework that improves generalization under non-IID and noisy label settings. FedRog introduces a parameter-aware selection and fine-tuning mechanism to align global and local representations, and a neighbor embedding consistency constraint to enhance robustness against noisy supervision. Furthermore, a fine-grained, importance-guided global aggregation strategy based on Fisher information is employed to mitigate unreliable updates from low-quality clients. We conduct extensive experiments on 16 graph classification datasets under five heterogeneous data partition settings. Results show that FedRog consistently achieves competitive or superior performance compared to 14 baselines in terms of both accuracy and robustness under clean and noisy conditions.

Mobile applications based on machine learning (ML) are increasingly relying on offloading to the edge devices for low-latency, resource-efficient computation. Applying serverless computing for these ML applications on the edge offers a promising solution for handling dynamic workloads while meeting user-specified latency service-level objectives (SLOs). However, existing serverless frameworks, with their coarse-grained data parallelism and rigid model partitioning, are inadequate for ML inference on widely adopted edge System-on-Chip (SoC) devices. This paper presents FluidEdge, an edge-native serverless inference framework. FluidEdge identifies bottleneck operators in ML models and addresses them through a novel fine-grained intra-function latency-sensitive auto-scaling approach that dynamically scales inference bottlenecks during online serving. Additionally, it employs inter-function scaling to further prevent latency SLO violations and leverages the unified memory of edge SoCs for efficient data sharing during inference. Experimental results demonstrate that FluidEdge achieves a 37.4% latency improvement and 67.3%-87.6% SLO violation reduction compared to best-performed state-of-the-art serverless inference frameworks.

As a new paradigm of machine learning, green learning has achieved performance comparable to deep learning in vision tasks, knowledge graph learning, and modeling graph structures. Compared with deep learning, green learning has the advantages of a low carbon footprint, lightweight model and logical transparency. In this paper, to enhance the robustness of green learning, we study the performance of the green model when the training dataset labels are noisy. Through theoretical analysis and experimental verification, we found that noisy labels will not only reduce the performance of green learning, but also amplify the risk of privacy leaks of training dataset members. Therefore, we propose a robust green learning to approach the above two threats. Specifically, we design a multi-stage label-consistency sample selection method to filter out noisy labels in the training dataset by taking advantage of the characteristics of unsupervised representation learning in green learning. In order to further improve the stable classification performance of the model, we propose a feature-level class-balanced data augmentation method in the feature decision learning stage to solve the class imbalance problem caused by sample selection. Finally, a large number of experiments on multiple noisy datasets show that the robust green learning method proposed in this paper can not only enhance the generalization of the model under noisy training labels, but also alleviate the risk of member privacy information leakage.