https://github.com/zacharyvarley/approxpcatorch

Oja and implicit Krasulina approaches to batched approximate top-k PCA estimation.
https://github.com/zacharyvarley/approxpcatorch

eigenface incremental-pca online-pca-algorithms pca pca-analysis pytorch pytorch-implementation stochastic-optimization subspace-decomposition subspace-learning

Last synced: about 1 month ago
JSON representation

Oja and implicit Krasulina approaches to batched approximate top-k PCA estimation.

• Host: GitHub
• URL: https://github.com/zacharyvarley/approxpcatorch
• Owner: ZacharyVarley
• License: mit
• Created: 2025-02-21T20:43:22.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-02-22T00:03:43.000Z (over 1 year ago)
• Last Synced: 2025-03-04T08:18:01.181Z (over 1 year ago)
• Topics: eigenface, incremental-pca, online-pca-algorithms, pca, pca-analysis, pytorch, pytorch-implementation, stochastic-optimization, subspace-decomposition, subspace-learning
• Language: Python
• Homepage:
• Size: 647 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# ApproxPCATorch

Oja and implicit Krasulina approaches to batched approximate top-k PCA estimation. See mnist_test.py for usage. Batch size, dimension, and k all heavily influence which method might be the best fit for a given application.

![PCA components for SVD and four approximations](mnist_kpca_components.png)