PCA on a 3d Matrix

Hi. I have a data set of http://www.ehu.eus/ccwintco/index.php/Hyperspectral_Remote_Sensing_Scenes Indian Pines. Data set is a 145x145x200 matrix that 145x145 represents spatial dimensions, 200 represents feature dimensions. I wnat to do PCA dimension reduction but I can't figure out how to deal 3d matrix. If you could help me I would be appreciated. Thanks.

[KALYAN ACHARJYA]

Have you checked here?

https://www.mathworks.com/matlabcentral/answers/410512-how-can-i-perform-a-pca-analysis-over-3d-data

[Micheal Dennington]

Yeah I saw that example but I couldn't understand that very much.

[Sanchay Mukherjee]

Hi Michael,

Did you figure out hte solution? I am trying to do a similar thing. I have a matrix of 200*500*3, where 200*500 is the data for corresponding 3 features.

Answer 1

[Subhadeep Koley]

You can use the hyperpca function to achieve the same.

% Definenumber of principal components you require
numComponents = 10;
% Perform PCA transform
outputDataCube = hyperpca(indianPinesMatrix, numComponents);

The above mentioned feture comes under Image Processing Toolbox's Hyperspectral Imaging Library support package, and can be downloaded from here. For more information on Hyperspectral Imaging Library see the documentation.