GOODAT: Towards Test-Time Graph Out-of-Distribution Detection

Authors: Luzhi Wang, Dongxiao He, He Zhang, Yixin Liu, Wenjie Wang, Shirui Pan, Di Jin, Tat-Seng Chua

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Comprehensive evaluations confirm that our GOODAT method outperforms state-of-the-art benchmarks across a variety of real-world datasets.
Researcher Affiliation Academia 1College of Intelligence and Computing, Tianjin University 2Faculty of Information Technology, Monash University 3School of Computing, National University of Singapore 4School of Information and Communication Technology, Griffith University
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement about open-sourcing the code for the described methodology or a link to a code repository.
Open Datasets Yes The graph OOD detection benchmark contains 8 pairs of molecule datasets, 1 pair of bioinformatics datasets, and 1 pair of social network datasets. ... The graph anomaly detection benchmark comprises 15 datasets from TU benchmark (Morris et al. 2020).
Dataset Splits No The paper mentions a 90% training and 10% test split, but does not explicitly provide details for a separate validation split or how it was handled.
Hardware Specification Yes Experiments run on a GeForce GTX TITAN X GPU with 24 GB memory
Software Dependencies No The paper mentions using the Adam optimizer but does not specify version numbers for any software components or libraries.
Experiment Setup Yes We use the Adam optimizer (Kingma and Ba 2014) for optimization. ... We conduct a parameter sensitivity experiment on the PTC-MR/MUTAG dataset, where α is selected from {0.1, 0.3, 0.5, 0.7, 0.9} and β is selected from {0.01, 0.03, 0.05, 0.07, 0.09}. As shown in Fig. 5 (a), when β is fixed, optimal outcomes are achieved with α in the range of 0.1-0.3. Likewise, when α is held constant, β values in the range of 0.3-0.5 yield optimal results...