Summary
Now you understand that there is a method to the analytics madness. You also now know that there are multiple approaches you can take to data science problems. You understand that building a model on captive data in your own machine is an entirely different process from deploying a model in a production environment. You also understand different approaches to the process and that you and your stakeholders may each show preferences for different ones. Whether you are starting with the data exploration or the problem statement, you can find useful and interesting insights.
You may also have had your first introduction to the overlays and underlays concepts, which are important concepts as you go deeper into the data that is available to you from your network in the next chapter. Getting data to and from other overlay applications, as well as to and from other layers of the network is an important part of building complete solutions.
You now have a generalized analytics infrastructure model that helps you understand how the parts of analytics solutions come together to form a use case. Further, you understand that using the analytics infrastructure model allows you to build many different levels of analytics and provides repeatable, reusable components. You can choose how mature you wish your solution to be, based on factors from your own environment. The next few chapters take a deep dive into understanding the networking data from that environment.