So far we have used logistic regression to create our machine learning model. We will now cover some alternative methods, starting with Support Vector Machines. We can keep nearly of our code. In fact we only have to change the three lines in the ’train mode’ function.
We’ll re-use the logistic regression code for looking for appropriate regularisation of the model (to avoid over-fitting of the model to the training data). Our three lines of code which define the model are now: Continue reading “71. Machine Learning: Support Vector Machines”