Hi! For those who are wondering why we left support vector machines (SVMs) out of the updated machine learning specialization. Andrew explained that he no longer uses SVMs anymore, as they’ve been superseded by neural networks. For binary classification problems, logistic regression is a relatively simple method compared to SVMs (and logistic regression is really just a neural network with a single neuron), so you may still prefer to use logistic regression for binary classification when speed matters. But if you’re planning to use a more complex algorithm, then Andrew would recommend that you just use a neural network instead of SVM.
Hope that helps those of you who were wondering about this. 