Identifying ML platform use cases
As discussed in the earlier chapters, it is imperative to understand what ML is and how it differs from other closely related disciplines, such as data analytics and data science. Data science may be required as a precursor to ML. It is instrumental in the research and exploration phase where you are unsure whether an ML algorithm can solve the problem. In the previous chapters, you have employed data science practices such as problem definitions, isolation of business metrics, and algorithm comparison. While data science is essential, there are also ML use cases that do not require as many data science activities. An example of such cases is the use of AutoML frameworks, which we will talk about in the next section.
Identifying whether ML can best solve the problem and selecting the ML platform is a bit of a chicken and egg problem. This is because, in order to be sure that an ML algorithm can best solve a certain business problem, it requires...
