Appendix B. The AI Process

This chapter covers an AI and machine learning process model that I created called the GABDO AI Process Model. GABDO models the end-to-end process of actually preparing for, developing, delivering, and improving specific AI and machine learning tasks and projects.

There are other variations of this process that you might encounter such as CRISP-DM and knowledge discovery in databases, aka KDD. The existing process models are great and widely used. That said, I find that the majority of resources and content on AI, machine learning, and data science is geared primarily toward practitioners such as data scientists and machine learning engineers. As a result, my motivation in creating the GABDO model isn’t to reinvent the wheel, but rather to reimagine the process in a way that I think is better suited to executives and managers.

With that, I introduce the GABDO AI Process Model. I refer only to AI in the model name because machine learning can be considered a subset of AI, so everything presented here applies to machine learning and other AI techniques.

The GABDO Model

Figure B-1 shows the GABDO AI Process Model.

The GABDO AI Process Model consists of five iterative phases—goals, acquire, build, deliver, optimize—hence, represented by the acronym GABDO. Each phase is iterative because any phase can loop back to one or more phases before. The typical duration of each phase is highly dependent on many factors, including the scientific, empirical, and nondeterministic ...

Get AI for People and Business now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.