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

https://github.com/thefloatingstring/cpca-as-preprocessing-for-supervised-models


https://github.com/thefloatingstring/cpca-as-preprocessing-for-supervised-models

Last synced: 5 months ago
JSON representation

Awesome Lists containing this project

README

          

# Improving Classification Accuracy using Contrastive Principal Component Analysis

Notebooks can be run using Google Colab. The Google Colab environment should automatically take care of all dependencies and data downloads.

### Notebooks

* `compare-cpca-with-other-methods.ipynb`: compares cPCA with other dimensionality reduction methods (runtime: ~34 seconds)
* `run-cpca-on-mnist-and-mouse-datasets.ipynb`: runs cPCA experiments on the MNIST and Mouse Down Syndrome Gene Expression datasets (runtime: ~244 seconds)
* `validate-cPCA-with-kNN.ipynb`: evalute k-NN 5-fold mean validation accuracy using PCA and cPCA pre-processed data (runtime: ~15 seconds)

### Runtime Instructions

For `run-cpca-on-mnist-and-mouse-datasets.ipynb`, the user would need to specify a JSON object containing a valid Kaggle username and password, as specified in the notebook instructions.

The other two notebooks can be run without any additional configuration.