Deep Adversarial Multi-view Clustering Network

Authors: Zhaoyang Li, Qianqian Wang, Zhiqiang Tao, Quanxue Gao, Zhaohua Yang

IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on several real-world datasets demonstrate the proposed method outperforms the state-of art methods. 4 Experiments
Researcher Affiliation Academia 1State Key Laboratory of Integrated Services Networks, Xidian University, Xi an 710071, China. 2Department of Electrical and Computer Engineering, Northeastern University, USA. 3School of Instrumentation Science and Opto-electronics Engineering, Beihang University, China.
Pseudocode No No pseudocode or algorithm blocks were found.
Open Source Code No The paper does not provide a link to open-source code or explicitly state that the code is available.
Open Datasets Yes Handwritten numerals (HW) dataset [Asuncion and Newman, 2007] is composed of 2,000 data points from 0 to 9 ten digit classes and each class has 200 data points. BDGP [Cai et al., 2012] is a twoview dataset including two different modalities, i.e., visual and textual data. The Columbia Consumer Video (CCV) dataset [Jiang et al., 2011] contains 9,317 You Tube videos with 20 diverse semantic categories. MNIST is a widely-used benchmark dataset consisting of handwritten digit images with 28 28 pixels. In our experiment, we employ its two-view version (70, 000 samples) provided by [Shang et al., 2017]
Dataset Splits No The paper does not explicitly provide details about training, validation, and test dataset splits (e.g., percentages or sample counts).
Hardware Specification Yes We run all the experiments on the platform of Ubuntu Linux 16.04 with NVIDIA Titan Xp Graphics Processing Units (GPUs) and 32 GB memory size.
Software Dependencies No The paper mentions "Py Torch" and "Adam optimizer" but does not specify version numbers for these software components.
Experiment Setup Yes We use Adam [Kingma and Ba, 2014] optimizer with default parameter setting to train our model and fix the learning rate as 0.0001. We conduct 30 epochs for each training step.