1 / 15

Principal Component Analysis (PCA)

Principal Component Analysis (PCA). Dated back to Pearson (1901) A set of data are summarized as a linear combination of an ortonormal set of vectors which define a new coordinate system. Figure from Hastie, et, al 2001. Principal Components Analysis (PCA). Principal Components Analysis (PCA).

kelly-frye
Download Presentation

Principal Component Analysis (PCA)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Principal Component Analysis (PCA) • Dated back to Pearson (1901) • A set of data are summarized as a linear combination of an ortonormal set of vectors which define a new coordinate system. Figure from Hastie, et, al 2001

  2. Principal Components Analysis (PCA)

  3. Principal Components Analysis (PCA)

  4. Singular Value Decomposition

  5. Singular Value Decomposition

  6. Singular Value Decomposition

  7. Singular Value Decomposition

  8. Example 1 • Use the data set "noisy.mat" available on the course web page. The data set consists of 1965, 20-pixel-by-28-pixel grey-scale images distorted by adding Gaussian noises to each pixel with s=25.

  9. Example 1 • Apply PCA to the noisy data. Suppose the intrinsic dimensionality of the data is 10. Compute reconstructed images using the top d = 10 eigenvectors and plot original and reconstructed images

  10. Example 1 • If original images are stored in matrix X (it is 560 by 1965 matrix) and reconstructed images are in matrix X_hat , you can type in • colormap gray and then • imagesc(reshape(X(:,10),2028)’) • imagesc(reshape(X_hat(:,10),2028)’) to plot the 10th original image and its reconstruction.

  11. Example 2

  12. Example 2 • Load the sample data, which includes digits 2 and 3 of 64 measurements on a sample of 400. load 2_3.mat • Extract appropriate features by PCA [u s v]=svd(X','econ'); • Create data Low_dimensional_data=u(:,1:2); • Observe low dimensional data Imagesc(Low_dimensional_data)

  13. Kernel Methods

  14. Kernel Trick

  15. Observed and Feature

More Related