Skip to content
Snippets Groups Projects
Select Git revision
  • f88030af4eb954b094578b35d8704376de73ec76
  • master default protected
2 results

02456-deep-learning-final-projektet

  • Open with
  • Download source code
  • Your workspaces

      A workspace is a virtual sandbox environment for your code in GitLab.

      No agents available to create workspaces. Please consult Workspaces documentation for troubleshooting.

  • user avatar
    Lin Manxi authored and GitHub committed
    f88030af
    History

    DIFF GRAPH U-NET: A DIFFERENTIABLE U-SHAPE NETWORK FOR GRAPH CLASSIFICATION

    Introduction

    Recently, graph convolutional networks (GCNs) have been demonstrated efficient in learning graph representations. Regarding the down-sampling and up-sampling of non-Euclidean data, most existing methods are flat and lack robustness. We visualize the process of a state-of-the-art work DiffPool, and develop a novel differentiable module for upsampling called DiffUnpool. DiffPool and DiffUnpool learn soft cluster assignment for nodes via GCNs and multi-layer perceptrons respectively. To address the graph classification problem, based on DiffPool and DiffUnpool, we further propose an end-to-end encoder-decoder architecture, diff graph U-Net. Different from other U-shape models before, diff graph U-Net learns node embeddings hierarchically, and collect global features in residual fashion. Our experimental results show that our model yields an overall improvement of accuracy on 4 different data sets, compared with previous methods.

    Check our report

    https://github.com/mmmmimic/02456-deep-learning-final-projektet/blob/master/report.pdf

    Author

    Manxi Lin s192230

    Mengge Hu s192113

    Guangya Shen s200104

    Data set

    https://drive.google.com/file/d/1nTM9c4HgIeb6iFauLQABuGjqDGpc43iv/view?usp=sharing

    Unzip it in here

    Check our main result

    • See main.ipynb
    • Proof of our result: ./screenshots

    Our contribution

    • Revision in encoders.py and train.py to implement DIFF Graph U-Net
    • Implement sample_data.py to sample data sets
    • Display our main result in a jupyter notebook main.ipynb

    DIFFPOOL

    Repo Link https://github.com/RexYing/diffpool

    DIFFPOOL Introduction (from the origional repo)

    Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings, and achieved state-of-the-art results in tasks such as node classification and link prediction. However, current GNN methods are inherently flat and do not learn hierarchical representations of graphs—a limitation that is especially problematic for the task of graph classification, where the goal is to predict the label associated with an entire graph. Here we propose DIFFPOOL, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. DIFFPOOL learns a differentiable soft cluster assignment for nodes at each layer of a deep GNN, mapping nodes to a set of clusters, which then form the coarsened input for the next GNN layer. Our experimental results show that combining existing GNN methods with DIFFPOOL yields an average improvement of 5–10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark data sets.

    Paper link: https://arxiv.org/pdf/1806.08804.pdf