QUESTION 1 Which of the following statements is wrong? When the loss function is convex, the gradient descent algorithm will find the optimal solution When performing PCA dimensionality reduction, we need to calculate the covariance matrix When performing the gradient descent algorithm, the direction along the negative gradient is the optimal direction Using the Lagrange multiplier can solve the optimization problem with constraints QUESTION 2 Which of the following statements about PCA is correct? The covariance matrix is S=N1XTX When using PCA, we want to maximize the variance of the projected data When using PCA, we want to maximize the reconstruction error When projecting data onto a k-dimensional space, the first k eigenvectors of the covariance matrix will be used Which of the following statements is not a dimensionality reduction method? PCA decision tree MDS F-score The centering matrix is H=INN11N1NT, where IN is an identity matrix and 1N=[1,1,,1]T. Which of the following statements is wrong? H=HTHn=Hmax(H)