Summary

We've seen how SVMs can be used for both classification and regression and learned a few basics about how they work. For classification, SVMs optimize a hyperplane to separate classes, maximizing the margin between the hyperplane and the datapoints. We can tune the performance with hyperparameters such as C and L1/L2 regularization. For regression, we can use the epsilon hyperparameter to balance the model's bias and variance. Finally, we saw how pycaret can be used to easily tune the SVM models and search hyperparameter spaces.

We've taken a look at several models for classification and regression for supervised learning, which is a major part of machine learning. However, if we don't have target values to predict, we can still use machine learning. In that case, we can use clustering, which helps us uncover patterns in data. In the next chapter, we'll cover some important clustering algorithms and how to use them.