Meta-Learned Metrics over Multi-Evolution Temporal Graphs

Fu, Dongqi; Fang, Liri; Maciejewski, Ross; Torvik, Vetle I.; He, Jingrui

doi:10.1145/3534678.3539313

Cited by 17 publications

(8 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Forward problems on graph diffusion. The vast majority of research on diffusion or dynamic graphs [29,30] are devoted to forward problems. Pioneering works derive epidemic thresholds for random graphs from probabilistic perspectives [6] or for arbitrary graphs from spectral perspectives [31,63,85,89].…”

Section: Related Workmentioning

confidence: 99%

Reconstructing Graph Diffusion History from a Single Snapshot

Qiu

Wang

Lü

et al. 2023

Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Diffusion on graphs is ubiquitous with numerous high-impact applications, ranging from the study of residential segregation in socioeconomics and activation cascading in neuroscience, to the modeling of disease contagion in epidemiology and malware spreading in cybersecurity. In these applications, complete diffusion histories play an essential role in terms of identifying dynamical patterns, reflecting on precaution actions, and forecasting intervention effects. Despite their importance, complete diffusion histories are rarely available and are highly challenging to reconstruct due to ill-posedness, explosive search space, and scarcity of training data. To date, few methods exist for diffusion history reconstruction. They are exclusively based on the maximum likelihood estimation (MLE) formulation and require to know true diffusion parameters. In this paper, we study an even harder problem, namely reconstructing Diffusion history from A single SnapsHot (DASH), where we seek to reconstruct the history from only the final snapshot without knowing true diffusion parameters. We start with theoretical analyses that reveal a fundamental limitation of the MLE formulation. We prove: (a) estimation error of diffusion parameters is unavoidable due to NP-hardness of diffusion parameter estimation, and (b) the MLE formulation is sensitive to estimation error of diffusion parameters. To overcome the inherent limitation of the MLE formulation, we propose a novel barycenter formulation: finding the barycenter of the posterior distribution of histories, which is provably stable against the estimation error of diffusion parameters. We further develop an effective solver named DIffusion hiTting Times with Optimal proposal (DITTO) by reducing the problem to estimating posterior expected hitting times via the Metropolis-Hastings Markov chain Monte Carlo method (M-H MCMC) and employing an unsupervised graph neural network to learn an optimal proposal to accelerate the convergence of M-H MCMC. We conduct extensive experiments to demonstrate the efficacy of the proposed method. Our code is available at https://github.com/q-rz/KDD23-DITTO.

show abstract

Section: Related Workmentioning

confidence: 99%

Reconstructing Graph Diffusion History from a Single Snapshot

Qiu

Wang

Lü

et al. 2023

Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

show abstract

“…For example, in Fu and He (2021a), each dynamic protein-protein interaction network has 36 continuous observations (i.e., 36 edge timestamps), every 12 observations compose a metabolic cycle (i.e., three snapshots), and each cycle reflects 25 mins in the real world. Inspired by this observation, a nascent work (Fu et al, 2022b) is recently proposed to jointly model different evolution patterns into the graph representation.…”

Section: Natural and Artificial Dynamics In Graphsmentioning

confidence: 99%

“…Generally speaking, if we model each evolution pattern as a different view of the input graph, then VANE (Fu et al, 2020d) could get the node embedding that is suitable for each observed view. Specifically, Temp-GFSM (Fu et al, 2022b) is proposed, which deliberately targets the streaming pattern for rapid node/edge-level evolution and the snapshot pattern for episodic and slowly-changing evolution, as shown in Figure 3. In Temp-GFSM, a multi-time attention mechanism is introduced with the support of the time kernel function to get the node-level, snapshot-level, and graph-level embeddings across different evolution patterns.…”

Section: Natural Dynamics In Graph Representationsmentioning

confidence: 99%

Natural and Artificial Dynamics in Graphs: Concept, Progress, and Future

2022

Front. Big Data

View full text Add to dashboard Cite

Graph structures have attracted much research attention for carrying complex relational information. Based on graphs, many algorithms and tools are proposed and developed for dealing with real-world tasks such as recommendation, fraud detection, molecule design, etc. In this paper, we first discuss three topics of graph research, i.e., graph mining, graph representations, and graph neural networks (GNNs). Then, we introduce the definitions of natural dynamics and artificial dynamics in graphs, and the related works of natural and artificial dynamics about how they boost the aforementioned graph research topics, where we also discuss the current limitation and future opportunities.

show abstract

“…Generic FSGC [220] [221], [222] Adaptive step controller [220], super-class graph [223], task correlations [221] Cross-domain FSGC --Data augmentation [224] Few-shot temporal graph classification [225] [225] -Few-shot molecular property prediction [226]- [229] -Meta-task reweighting [227], implicit function theorem [230] Tasks Other techniques TABLE 5: Summary of few-shot graph classification.…”

Section: Maml Prototypical Networkmentioning

confidence: 99%

“…Few-shot temporal graph classification. Fu et al [225] explore a novel setting of few-shot temporal graph classification, where each class may have only a few labeled temporal graphs and novel classes might emerge in the future. They present Temp-GFSM [225], a temporal graph metric learning framework.…”

Section: Few-shot Graph Classificationmentioning

confidence: 99%