[Solved] In R Studio: The pendigits dataset consists of handwritten digits…

In R Studio:

The pendigits dataset consists of handwritten digits (see Section 7.2.10 for a
full description). From this set, extract the subset corresponding to digits 0,
6, 8 and 9, and scale the variables so that the variances are 1.
(a) Compute ordinary principal components and draw a scatterplot of the
first two component scores, using different colors (or symbols) for different
digits. Are the digits well separated?
(b) Now compute kernel principal components, using a Gaussian kernel, and
draw a scatterplot of the first two component scores as in (a). Are the
digits better separated than in (a)?

 

Read more here: Source link