Problem
For this problem, you will consider the ENTIRE Iris 2 class data frame consisting of both setosas and versicolors. First, create TWO additional predictor variables consisting of (i) the square root of sepal width (Hint: this can be computed using np.sqrt(df.sepal_width)) and (ii) the square of the sepal width. THEN drop the 2 predictors petal length and sepal width. Also drop the actual class label. Now use the numerical CLASS LABEL (0 or 1) as the target (i.e., the predicted variable) and all the other predictors in the data frame used to predict the target. Build an SVM model to predict one of the 2 classes, setosas or versicolors. Use the entire data set to build the SVM (no training and test sets).
What is the predicted class for a NEW record X_New as below?
X_new = [[1.6,0.2,1.27,2.56]]
(=