In this chapter, we have explored SVMs and how they can be used to fit both linear and nonlinear data. The following are the other topics that we have covered:
We have examined how SVMs are capable of large margin classification and the various forms of the optimization problem of SVMs
We have discussed how we can use kernel functions and SMO to train an SVM with nonlinear sample data
We have also demonstrated how we can use several Clojure libraries to build and train SVMs
We will shift our focus to unsupervised learning in the next chapter and we will explore clustering techniques to model these types of machine learning problems.