In this chapter, we had a brief introduction to the topic and got a grasp of simple, yet powerful and common ML techniques. Finally, you saw how to build your own predictive model using Spark. You learned how to build a classification model, how to use the model to make predictions, and finally, how to use common ML techniques such as dimensionality reduction and One-Hot Encoding.

In the later sections, you saw how to apply the regression technique to high-dimensional datasets. Then, you saw how to apply a binary and multiclass classification algorithm for predictive analytics. Finally, you saw how to achieve outstanding classification accuracy using a random forest algorithm. However, we have other topics in machine learning that need to be covered too, for example, recommendation systems and model tuning for even more stable performance before you finally deploy the models.

In the next chapter, we will cover some advanced topics of Spark. We will provide examples of machine learning model tuning for better performance, and we will also cover two examples for movie recommendation and text clustering, respectively.