/Clustering-on-the-MNIST-data

Comparisons between KMeans, PCA, LDA, t-SNE on the MNIST dataset

Primary LanguageJupyter Notebook

Clustering-on-the-MNIST-data

Intro:

MNIST is a well known handwritten digits dataset intended for image classification. Currently there are multiple popular dimension reduction and classification algorithms and a comparison has been made between KMeans, PCA, LDA, t-SNE on the MNIST dataset.

Demo:

Eigenvalues:

demo

Kmeans:

demo

LDA:

demo

PCA:

demo

t-SNE

demo

http://woodenleaves.com/pages/cluster.html