(E) Least squares regression with appropriate response variables

The misclassification rate for Least square regression for above data comes out to 0.496 for the test data and 0.45 for the train data.

(F) SVM

The misclassification for SVM Linear comes out to be 0.44 for the test and 0.39 for the train data.
The misclassification for SVM Polynomial comes out to be 0.356 for the test and 0.36 for the train data.
Using SVM Radial gives 0.29 misclassification rate for the test data and 0.344 for the train data.