Link prediction is a fundamental problem in network modeling. A family of link prediction approaches is to treat network data as an exchangeable array whose entries can be explained by random functions (e.g., block models and Gaussian processes) over latent node factors. Despite their powerful ability in modeling missing links, these models tend to have a large computational complexity and thus are hard to deal with large networks. To address this problem, we develop a novel variational random function model by defining latent Gaussian processes on exchangeable arrays. This model not only inherits the ability of Gaussian process to describe the nonlinear interactions between nodes, but also enjoys significant reduction on computational complexity. To further make the model scalable to large network data, we develop an efficient key-value-free strategy under the map-reduce framework to tremendously reduce the inference time. Experimental results on large network data have demonstrated both the efficacy and efficiency of the proposed method over state-of-the-arts methods in network modeling.
This study explores the dynamics of industrial value chains from the perspective of directed graph neural networks (DGNNs). This study focuses on the effects of environmental, social, and governance (ESG) factors on industrial value extension. The industrial value chain, which encompasses a comprehensive network ranging from raw material procurement to final product distribution, is characterized by intricate interconnections between internal and external stakeholders. Traditional quantitative methods, such as input–output (I–O) analysis, often fail to capture the complexity of these relationships. We model the shock propagation across the industrial chain by employing DGNNs, defining two types of hidden representations: demanding (incoming) and supplying (outgoing) embeddings. This dual representation enables a nuanced understanding of how ESG shocks influence the extension of industrial value from downstream and upstream industrial sectors. Our findings highlight potential areas for optimization and value-added opportunities, offering insights that can support more informed decision-making by corporations and policymakers as they navigate the evolving industrial landscape.
24 March 2024 in Proceedings of the AAAI Conference on Artificial Intelligence
Existing semi-supervised domain adaptation (SSDA) models have exhibited impressive performance on the target domain by effectively utilizing few labeled target samples per class (e.g., 3 samples per class). To guarantee an equal number of labeled target samples for each class, however, they require domain experts to manually recognize a considerable amount of the unlabeled target data. Moreover, as the target samples are not equally informative for shaping the decision boundaries of the learning models, it is crucial to select the most informative target samples for labeling, which is, however, impossible for human selectors. As a remedy, we propose an EFfective Target Labeling (EFTL) framework that harnesses active learning and pseudo-labeling strategies to automatically select some informative target samples to annotate. Concretely, we introduce a novel sample query strategy, called non-maximal degree node suppression (NDNS), that iteratively performs maximal degree node query and non-maximal degree node removal to select representative and diverse target samples for labeling. To learn target-specific characteristics, we propose a novel pseudo-labeling strategy that attempts to label low-confidence target samples accurately via clustering consistency (CC), and then inject information of the model uncertainty into our query process. CC enhances the utilization of the annotation budget and increases the number of “labeled” target samples while requiring no additional manual effort. Our proposed EFTL framework can be easily coupled with existing SSDA models, showing significant improvements on three benchmarks