Homework 5

• Given two classes of data on a plane: (0,0), (0,2), (1,1), (2,0), (2,2) are in the first class and (0,1), (1,0), (1,2), (2,1) are in the second. We would like to have a regression model y = w_1 x_1 + w_2 x_2 using y_i = 1 for the first class and y_i = -1 for the second class. After the model is obtained, calculate training error by the following way: mean-squared error, mean absolute error, root relative squared error, relative absolute error, correlation coefficient.

• Consider the same arff file used in homework 3. Now we would like to use C4.5 implementation in weka and see what's the best result by 5-fold cross validation. Note you have to adjust the parameters.

  Now the test file is also available here. (compressed by bzip2) Using your best setting from cross valiation, report your test accuracy.