An open API service indexing awesome lists of open source software.

https://github.com/borgwardtlab/topo-ae-distances


https://github.com/borgwardtlab/topo-ae-distances

Last synced: 3 months ago
JSON representation

Awesome Lists containing this project

README

          

# Challenging Euclidean Topological Autoencoders

This is a follow-up project of the ICML 2020 paper "Topological Autoencoders" (reference below).
Here, we investigate whether domain-specific distance functions in the input space (here image datasets) are necessary for TopoAE, or whether a generic euclidean distance is sufficient.
[This work](https://openreview.net/pdf?id=P3dZuOUnyEY) has been accepted for presentation at the Neurips 2020 TDA and Beyond workshop.

## References

Please use the following BibTex code to cite our Neurips 2020 workshop [paper](https://openreview.net/pdf?id=P3dZuOUnyEY):

```
@InProceedings{moor2020challenging,
title = {Challenging Euclidean Topological Autoencoders},
author = {Moor, Michael and Horn, Max and Borgwardt, Karsten and Rieck, Bastian},
booktitle = {NeurIPS 2020 Workshop on Topological Data Analysis and Beyond},
year = {2020},
url = {https://openreview.net/forum?id=P3dZuOUnyEY},
}
```

Furthermore, the original ICML 2020 [paper](https://arxiv.org/abs/1906.00722) proposing Topological Autoencoders in the first place, can be cited as follows:

```
@InProceedings{Moor19Topological,
author = {Moor, Michael and Horn, Max and Rieck, Bastian and Borgwardt, Karsten},
title = {Topological Autoencoders},
year = {2020},
eprint = {1906.00722},
archiveprefix = {arXiv},
primaryclass = {cs.LG},
booktitle = {Proceedings of the 37th International Conference on Machine Learning~(ICML)},
series = {Proceedings of Machine Learning Research},
publisher = {PMLR},
pubstate = {forthcoming},
}
```

## Setup
In order to reproduce the results indicated in the workshop paper simply setup an
environment using poetry:
```bash
poetry install
```

## Running the methods:
Make sure you have internet access once to be able to download the datasets, and also the vgg model (via the lpips package)

In case a slurm cluster is available, simply run:
```bash
source scripts/run_slurm.sh
```

Alternatively, all jobs can be sequentially/manually called using:
```bash
source scripts/run_manual.sh
```