Figure 2: Multiclass Logistic Regression (a) (5 points) Consider a multi-class classification problem with 5 classes and 20 features. We will use the logistic regression model to first build our binary classifier. Then, what will be the total number of parameters for using One vs. Rest (ovr) strategies for the multi-class classification task using logistic regression? (Hint. Refer to lecture 11 for multi-class logistic regression model.) (b) (5 points) Consider a new multi-class classification problem with 3 classes. The distribution of the points is shown in the figure (Square - Class 1, Circle - Class 2, Star - Class 3). Draw the linear classifiers used for classifying the three classes, using (i) One vs. Rest (OvR), and (ii) Multinomial approaches. Your drawing does not have to be precise, roughly indicate the classifiers and jut make sure you can show the differences.