Starting from:

$25

AMSC808N - Homework 5 - Solved

1.    (5 pts) Show that the Laplacian eigenmap to Rm is the solution to the following optimization problem:

                                      min∑kij∥yi − yj∥22 subject to             Y ⊺QY = I,          Y ⊺Q1m×1 = 0.                     (1)

i,j

Here, yi’s are columns of Y , and Y is m × m, and the rest of notation as in Section 7.3 of 4-DimReduction.pdf.

2.    (20 pts) The goal of this problem is to practice and compare various methods for dimensional reduction.

•    Methods:

(a)    PCA;

(b)    Isomap;

(c)    LLE;

(d)    t-SNE;

(e)    Diffusion map.

Diffusion map should be programmed from scratch. Readily available codes can be used for the rest. For example, the built-in Matlab function can be used for tSNE; S. Roweis’s code can be used for LLE; my code for isomap is in the lecture notes. If you use some standard code, specify its source, read its description, and be ready to adjust parameters in it.

•    Dataset 1: Scurve generated by MakeScurveData.m: 352 data points in 3D forming a uniform grid on the manifold.

 

Figure 1: Scurve

•    Dataset 2: Scurve generated by MakeScurveData.m and perturbed by Gaussian noise. Try various intensities, push each method to its limit.

•    Dataset 3: “Emoji” dataset generated by MakeEmojiData.m: a set of 1024 images each one is 40 × 40 pixels. Images vary from a smiley face to an angry face and in the degree of blurring. Its subsampled set is shown in Fig. 2. Note

1

that picking a good value of  for the diffusion map might require some effort as the distances between the nearest neighbors are very nonuniform. You should be able to get a nice 2D surface embedded into 3D with a right . Using α = 0 or α = 1 is up to you.


More products