In this exercise, you will investigate the use of a k-NN classifier on the Stock Market data used in the lab lesson, and in particular you will investigate the bias/variance trade-off behind the choice of k.
• Split the Stock Market data into a train and test partition, using the same criterion adopted in the lab lesson (temporal split).
• Using Lag1 and Lag2 as predictors, fit a k-NN classifier with different values of k (for example, from 1 to 100).
• Evaluate the train and test errors of each model and plot them as a function of 1/k.