Exercises

We have already given a lengthy code example of a simple kernelized algorithm in the main part of this chapter. Therefore, for the exercises, we will demonstrate some of the more complex aspects of kernel methods. Due to this increase in complexity, we will only ask a single question. It is intentionally challenging, so don’t be surprised if you don’t manage to complete it fully. Have a go at answering the exercise and then compare your answer to the one in the Answers_to_Exercises_Chap12.ipynb Jupyter notebook in the GitHub repository.

• The data in the kernel_PCA_matrix_data.csv file in the Data directory of the GitHub repository contains the matrix elements of N=200, 4x4 matrices. Each matrix corresponds to a single row of the .csv file. The column headings are of the form i_j, where i and j are integers, representing the i,j matrix element. For example, the column with the 1_3 heading holds the 1,3 matrix elements of each matrix. Use the data to perform...