to lasso and ridge regression methods to reconstruct an image from a set of corrupted data. The dataset contains of 2 txt file, hw1_Q5 X. txt and hw1_Q5_Y. txt. The original image is shown as the following: original image The dimension of the figure is 128 x 128 = 16384, but we only have 2304 of them avail- able. In class, this is identified as the "high-dimensional" problem: we are using 2304 samples to predict the value of 16384. We will experiment with different values of 1 E {0.000001, 0.0001, 0.01, 0.1, 1}. (a) (10 Points) Using the 5 penalty values of A mentioned above to fit the ridge regression model. You will have the coefficient B E R16384. Reshape B into 128 x 128 matrix, and plot the matrix as in the original image. Which coefficient yields the best result? (b) (10 Points) Using the 5 penalty values of A mentioned above to fit the lasso model. You will have the coefficient B E R16384. Reshape B into 128 x 128 matrix, and plot the matrix as in the original image. Which coefficient yields the best result? (c) (5 Points) Comparing the best results from lasso and ridge regression respectively, what