Two Class Classification Problem Two Features X X1 X2 Measured Training Data Classes Follo Q43828338
We have a two-class classification problem, where two features x=(x1,x2)’ are measured. The training data for these classes are as follows. Class A: {(3,2), (4,3),(5,2)} Class B: {(2,4),(1,4),(2,5)} First, a linear classifier f(x) = w’x+b is used. (a) (10 points] Give a sample linear classifier that goes through origin (x1,x2)=(0,0), by drawing the decision boundary and indicating the values of w and b. Second, SVM classifier f(x) = w’x+b that maximizes the margin between two classes is used. (b) (5 points] Draw the decision boundary. (c) (5 points] Indicate the support vectors. (d) (10 points) Determine the values of w and b. Show transcribed image text We have a two-class classification problem, where two features x=(x1,x2)’ are measured. The training data for these classes are as follows. Class A: {(3,2), (4,3),(5,2)} Class B: {(2,4),(1,4),(2,5)} First, a linear classifier f(x) = w’x+b is used. (a) (10 points] Give a sample linear classifier that goes through origin (x1,x2)=(0,0), by drawing the decision boundary and indicating the values of w and b. Second, SVM classifier f(x) = w’x+b that maximizes the margin between two classes is used. (b) (5 points] Draw the decision boundary. (c) (5 points] Indicate the support vectors. (d) (10 points) Determine the values of w and b.
Expert Answer
Answer to We have a two-class classification problem, where two features x=(x1,x2)’ are measured. The training data for these clas…
OR