Low-Code AI

Chapter 3. Machine Learning Libraries and Frameworks

This chapter introduces machine learning (ML) frameworks that simplify the development of ML models. Typically, you need to understand the underlying working principles of mathematics, statistics, and ML to build and train ML pipelines. These frameworks help you by automating many of the time-consuming ML workflow tasks such as feature selection, algorithm selection, code writing, pipeline development, performance tuning, and model deployment.

No-Code AutoML

Imagine you are a business analyst working for a utility company. You have a project that requires you to help the company develop marketing and outreach programs that target communities with high electrical energy consumption. The data is in a comma separated value (CSV) file format.

You do not have an ML background or any programming knowledge—but the team lead has asked you to take on this project because you have expressed an interest in ML and how it can be applied in the organization. Although you have no coding experience, the little research you have done has yielded a few observations:

• For noncoders like yourself, there are automated no-code ML frameworks with a graphical user interface (GUI) that you can use to build and train an ML model without writing a single line of code.

• For light coders, there are low-code ML frameworks that provide the ability to build and train an ML model by writing just a little bit of code.

• For seasoned coders, there are ML libraries ...