This repository contains the source code for the paper: An, Sungtae, et al. "ViVA: Semi-supervised Visualization via Variational Autoencoders." 2020 IEEE International Conference on Data Mining (ICDM). IEEE, 2020.
- Create a (conda) virtual environment and install the requirements
conda env create -f environment.yml
- Create a partially labeled dataset (MNIST)
python mnist_create_semisupervised_dataset.py --n-labeled 1000
- Train ViVA
python mnist_train_viva.py --dataset Data/mnist_1000_labeled.pkl
- Visualize the embeddings
python mnist_visualize_embeddings.py --model Results/best_checkpoint.pth