English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recently, the paradigm of pre-training and fine-tuning graph neural networkshas been intensively studied and applied in a wide range of graph mining tasks.Its success is generally attributed to the structural consistency betweenpre-training and downstream datasets, which, however, does not hold in manyreal-world scenarios. Existing works have shown that the structural divergencebetween pre-training and downstream graphs significantly limits thetransferability when using the vanilla fine-tuning strategy. This divergenceleads to model overfitting on pre-training graphs and causes difficulties incapturing the structural properties of the downstream graphs. In this paper, weidentify the fundamental cause of structural divergence as the discrepancy ofgenerative patterns between the pre-training and downstream graphs.Furthermore, we propose G-Tuning to preserve the generative patterns ofdownstream graphs. Given a downstream graph G, the core idea is to tune thepre-trained GNN so that it can reconstruct the generative patterns of G, thegraphon W. However, the exact reconstruction of a graphon is known to becomputationally expensive. To overcome this challenge, we provide a theoreticalanalysis that establishes the existence of a set of alternative graphons calledgraphon bases for any given graphon. By utilizing a linear combination of thesegraphon bases, we can efficiently approximate W. This theoretical finding formsthe basis of our proposed model, as it enables effective learning of thegraphon bases and their associated coefficients. Compared with existingalgorithms, G-Tuning demonstrates an average improvement of 0.5% and 2.6% onin-domain and out-of-domain transfer learning experiments, respectively.

Fine-tuning Graph Neural Networks by Preserving Graph Generative Patterns