Graphs are continually growing in size, and processing complex queries, such as multi-hop pattern queries, on them is becoming increasingly important. Although GPUs have received significant attention recently, there is still a notable shortage of efficient GPU-based out-of-memory methods for handling these queries. Three key issues arise when processing multi-hop queries on large-scale graphs using GPUs: the need for an efficient graph format, effective scheduling of accesses to graph partitions on storage, and dynamic buffer management on both the host and GPUs. To address these issues, we propose an efficient GPU-based out-of-memory multi-hop query processing framework called GFlux. Through extensive experiments, we have demonstrated that GFlux significantly improves both the speed and scalability compared to existing state-of-the-art methods.

The proliferation of mobile and Internet of Things (IoT) devices has resulted in a surge of time-series sensor data, posing significant challenges for centralized data collection and processing. This challenge has driven the adoption of edge computing, which offloads data processing to mid-level servers located at the edge of the Internet, thereby reducing computation and bandwidth demands. Federated learning has emerged as a promising method for training models in edge-computing environments. Recently, spatio-temporal graph neural networks (STGNNs) have shown impressive performance in time-series prediction, yet their application in edge computing is limited by the complexity of adapting them to distributed environments. To address this gap, we propose FedSTGNN (Federated Spatio-Temporal Graph Neural Network), a universal framework that converts existing centralized STGNN models into a federated learning version. We formulate the common STGNN training process using matrix operations, employ graph-based imputation methods to handle missing sensor values at edge servers, and facilitate the transition from centralized to federated STGNNs. Our comprehensive evaluations demonstrate that FedSTGNN not only preserves the prediction accuracy of the original STGNN models but is also significantly more network-efficient than the competing model. Furthermore, the framework proves its robustness in challenging real-world scenarios, including sparse graphs, long-term forecasting, and dynamic server participation. Our work presents a practical, robust, and universal solution for deploying STGNNs into various edge computing applications. • A framework for spatio-temporal graph neural network models in federated settings. • A matrix-based approach to convert STGNN training processes to federated settings. • Graph-based imputation methods compensate for missing data on edge servers. • Reduction of prediction error gap to those of original models on real-world datasets. • Reduction of training time and network usage, proving its practicality.

Deep neural networks are trained so as to achieve a kind of the maximum overall accuracy through a learning process using given training data. Therefore, it is difficult to fix them to improve the accuracies of specific problematic classes or classes of interest that may be valuable to some users or applications. To address this issue, we propose the synaptic join method to tweak neural networks by adding certain additional synapses from the intermediate hidden layers to the output layer across layers and additionally training only these synapses, if necessary. To select the most effective synapses, the synaptic join method evaluates the performance of all the possible candidate synapses between the hidden neurons and output neurons based on the distribution of all the possible proper weights. The experimental results show that the proposed method can effectively improve the accuracies of specific classes in a controllable way.