BIASLab-UCI / hd-clustering

Smart Hyperdimensional Clustering algorithm: FebHD. Mirror of https://gitlab.com/biaslab/hd-clustering

hd-clustering

Authors: Alejandro Hernández Cano, Mohsen Imani.

Installation

In order to install the package, simply run the following:

pip install febhd-clustering

Visit the PyPI project page for more information about releases.

Documentation

Read the documentation of this project.

Quick start

The following code generates dummy data and trains a FebHD clustering model with it.

>>> import febhd_clustering
>>> dim = 10000
>>> n_samples = 1000
>>> features = 100
>>> clusters = 5
>>> x = torch.randn(n_samples, features) # dummy data
>>> model = febhd_clustering.FebHD(clusters, features, dim=dim)
>>> if torch.cuda.is_available():
...     print('Training on GPU!')
...     model = model.to('cuda')
...     x = x.to('cuda')
...
Training on GPU!
>>> model.fit(x, epochs=10)
>>> ypred = model(x)
>>> ypred.size()
torch.Size([1000])

For more examples, see the examples/ directory.

Citation request

If you use hd-clustering, please cite the following papers:

1. Alejandro Hernández-Cano, Yeseong Kim, Mohsen Imani. "A Framework for Efficient and Binary Clustering in High-Dimensional Space". IEEE/ACM Design Automation and Test in Europe Conference (DATE), 2021.

2. Mohsen Imani, et al. "DUAL: Acceleration of Clustering Algorithms using Digital-based Processing In-Memory"r IEEE/ACM International Symposium on Microarchitecture (MICRO), 2020.