Update SupportVectorMachinesQuiz.md

Stanford_Machine_Learning/Week7/SupportVectorMachinesQuiz.md

@@ -57,3 +57,4 @@ True | The maximum value of the Gaussian kernel (i.e., sim(x,l<sup>(1)</sup>)) i
 True | Suppose you had 2D input examples (i.e. x<sup>(i)</sup> ∈ R<sup>2</sup>). The decision boundary of the SVM (with the linear kernel) is a straight line. | The SVM without any kernel (i.e., the linear kernel) predicts output based only Θ<sup>T</sup>x, so it gives a linear/straight-line decision boundary, just as logisitic regression does. 
 False | If you are training multi-class SVMs with the one-vs-all method, it is not possible to use a kernel | Each SVM you train in the one-vs-all method is a standard SVM, so you are free to use a kernel. 
 False | Suppose you are using SVMs to do multi-class classification and would likely to use the one-vs-all approach. If you have K different classes, you will train K-1 different SVMs. | The one-vs-all method requires that we have a separate classifier for every class, so you will train K different SVMs.
+True | It is important to perform feature normalization before using the Gaussian kernel. | The similarity measure used by the Gaussian kernel expects that the data lie in approximately the same range.