Incomplete Multi-view Clustering via Prototype-based Imputation (IJCAI2023)

This repo contains the code and data of our IJCAI'2023 paper Incomplete Multi-view Clustering via Prototype-based Imputation.

Incomplete Multi-view Clustering via Prototype-based Imputation

Requirements

pytorch>=1.2.0

numpy>=1.19.1

scikit-learn>=0.23.2

munkres>=1.1.4

Configuration

The hyper-parameters, the training options (including the missing rate) are defined in configure.py.

Datasets

The Scene-15, CUB datasets are placed in "data" folder.

Usage

The code includes:

• an example implementation of the model,
• an example clustering task for different missing rates.

python run.py --dataset 0 --devices 0 --print_num 50 --test_time 5

You can get the following output:

- Epoch : 50/150 ===> Learing Rate = 0.0010===> Ins loss = 1.6216e+01 ===> Clu loss = 0.0000e+00 ===> Loss = 1.6216e+01
-0.014516947
view_concat {'kmeans': {'AMI': 0.4169, 'NMI': 0.4217, 'ARI': 0.2382, 'accuracy': 0.3998, 'precision': 0.39, 'recall': 0.4057, 'f_measure': 0.3856}}
- Epoch : 100/150 ===> Learing Rate = 0.0010===> Ins loss = 1.5952e+01 ===> Clu loss = 8.9596e+00 ===> Loss = 2.4912e+01
-0.0060014944
- view_concat {'kmeans': {'AMI': 0.4248, 'NMI': 0.4295, 'ARI': 0.2565, 'accuracy': 0.4261, 'precision': 0.4072, 'recall': 0.428, 'f_measure': 0.407}}
- Epoch : 150/150 ===> Learing Rate = 0.0010===> Ins loss = 1.5866e+01 ===> Clu loss = 8.9524e+00 ===> Loss = 2.4818e+01
0.0018155271
- view_concat {'kmeans': {'AMI': 0.4321, 'NMI': 0.4368, 'ARI': 0.2705, 'accuracy': 0.443, 'precision': 0.4272, 'recall': 0.4439, 'f_measure': 0.4235}}

Reference

If you find our work useful in your research, please consider citing:

@inproceedings{li2023incomplete,
  title={Incomplete Multi-view Clustering via Prototype-based Imputation},
  author={Li, Haobin and Li, Yunfan and Yang, Mouxing and Hu, Peng and Peng, Dezhong and Peng, Xi},
  booktitle={Proceedings of the 32th International Joint Conference on Artificial Intelligence},
  year={2023}
}