24am

Parallel Sparse PCA Codes

Project has been moved to http://optml.github.io/24am

The source can be obtained here https://github.com/optml/24am

24 parallel sparse PCA codes

The library supports 8 optimization formulations of Sparse PCA:
L1/L2 variance x L0/L1 sparsity inducing norm x norm used as penalty/constraint

It contains parallel codes for the following architectures

  • multi-core CPUs (C++, Intel MKL, OpenMP)
  • GPUs (CUDA)
  • computer clusters (C++, BLASC)

Suitable for big data PCA and sparse PCA.

For more info please see our Wiki.


Based on the paper:

Peter Richtarik, Martin Takac and Selin D. Ahipasaoglu, http://arxiv.org/abs/1212.4137'>Alternating Maximization: Unifying Framework for 8 Sparse PCA Formulations and Efficient Parallel Codes, arXiv:1212.4137, December 2012

Project Information

The project was created on Sep 5, 2012.

Labels:
SparsePCA PCA Parallel ParallelPCA GPU Cluster