(Matlab) Codes and Datasets for Subspace Learning (Dimensionality Reduction)

Other standard data sets in matlab format

Algorithms

(Matlab 7 or higher version is required to open these files)

If you used the processed data sets on this page, we appreciate it very much if you can cite our following works:

Deng Cai, Xiaofei He, Jiawei Han, Thomas Huang, "Graph Regularized Non-negative Matrix Factorization for Data Representation", PAMI 2011.
Bibtex source
Deng Cai, Xiaofei He, Jiawei Han, "Speed Up Kernel Discriminant Analysis", The VLDB Journal, 2011.
Bibtex source

USPS database
The USPS handwritten digit database. We provide here a popular subset contains 9298 16x16 handwritten digit images in total, which is then split into 7291 training images and 2007 test images.
Data File: contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label. Sample images
The first 7291 images are training images and the last 2007 images are test images.
MNIST database
The MNIST database of handwritten digits from Yann LeCun's page has a training set of 60,000 examples, and a test set of 10,000 examples.
We provide here serveral smaller subsets. Sample images
The first 2k training images and first 2k test images: contains variables 'fea', 'gnd', 'trainIdx' and 'testIdx'. Each row of 'fea' is a sample; 'gnd' is the label.
The first 10k training images: contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label.
The test images (10k images in total): contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label.
All original images (the first 60k are the training images and the last 10k are the test images): contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label.
COIL20 database
It contains 20 objects. The images of each objects were taken 5 degrees apart as the object is rotated on a turntable and each object has 72 images. The size of each image is 32x 32 pixels, with 256 grey levels per pixel. Thus, each image is represented by a 1024-dimensional vector. Sample images
Data File: contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label.
COIL100 database
It contains 100 objects. The images of each objects were taken 5 degrees apart as the object is rotated on a turntable and each object has 72 images. The size of each image is 32x 32 pixels, with 256 grey levels per pixel. Thus, each image is represented by a 1024-dimensional vector. Sample images
Data File: contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label.
Isolet spoken letter recognition database
It contains 150 subjects who spoke the name of each letter of the alphabet twice. The speakers are grouped into sets of 30 speakers each, and are referred to as isolet1 through isolet5.
Isolet1: contains variables 'fea' and 'gnd'. Each row of 'fea' is a sample; 'gnd' is the label.
Isolet2
Isolet3
Isolet4
Isolet5

Return to Codes and Data