Problem 6: The training data has an accuracy of 90% while the test data only has 50% accuracy. What can you infer from this machine learning model? How can you improve it? (5 marks) Problem 7: You are given a feature vector X, a parameter vector W and a label Y. X=[1,x1,x2,x3]=[1,4,9,5]W=[w0,w1,w2,w3]=[0.5,0.8,1.0,0.3]Y=2 Gradient descent is an iterative optimization algorithm used to find the minimum of a loss function. Perform an iteration of gradient descent to find the updated values of the parameters. Assume learning rate, alpha, is 0.01. \{Hint: find predicted value of y and use residual sum of squares to find loss/error. Update values of W using gradient descent, and verify that the loss has decreased with the new set of parameters.) (10 marks)