Multi-Graph-View Learning for Complicated Object Classification

Authors: Jia Wu, Shirui Pan, Xingquan Zhu, Zhihua Cai, Chengqi Zhang

IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on real-world learning tasks demonstrate the performance of MGVBL for complicated object classification.
Researcher Affiliation Academia Quantum Computation & Intelligent Systems Centre, University of Technology, Sydney, Australia Dept. of Computer & Electrical Engineering and Computer Science, Florida Atlantic University, USA Dept. of Computer Science, China University of Geosciences, Wuhan 430074, China
Pseudocode Yes Algorithm 1 Discriminative Subgraph Exploration and Algorithm 2 MGVBL: Multi-Graph-View Bag Learning
Open Source Code No The paper does not provide any statement or link indicating the availability of open-source code for the described methodology.
Open Datasets Yes The Digital Bibliography & Library Project (DBLP) data set 1 consists of bibliography in computer science. 1http://dblp.uni-trier.de/xml/ and The original images [Li and Wang, 2008] from Corel data set2 are preprocessed by using VLFeat segmentation [Vedaldi and Fulkerson, 2008], with each image being segmented into multiple regions and each region corresponding to one graph. 2https://sites.google.com/site/dctresearch/Home/content-basedimage-retrieval
Dataset Splits Yes All reported results are based on 10 times 10-fold crossvalidation.
Hardware Specification No The paper does not provide any specific details regarding the hardware (e.g., GPU/CPU models, memory, or cloud resources) used to run the experiments.
Software Dependencies No The paper mentions software tools like VLFeat, SLIC, and gSpan, but does not specify their version numbers or list any other software dependencies with version information.
Experiment Setup Yes Unless specified otherwise, we set minimum support threshold min sup = 3% for scientific publication data (Section 4.3) and min sup = 2% for content-based image retrieval (Section 4.4). we set ϵ = 0.05 in our experiments.