Machine learning (ML), a subset of Artifical Intelligence (AI) is a fast growing area of the data sciences.

I had used certain Python libraries locally on a cloud based virtual machine on AWS, but wanted a overarching, detailed process for building ML applications.

I chose this book as it covers ML fundamentals through an end-to-end example project using Scikit-Learn and pandas. It also covers TensorFlow2 which was mentioned in the Google Cloud Architect Professional Certification exam.

What I found particularly interesting was which libraries perform what kind of analysis. So classification and regression, discover object detection, semantic segmentation, attention mechanisms, language models GANs and more.

It covers the Keras API, the official high-level API for TensorFlow 2.

Through this book I also learned how to productionize TensorFlow models using TensorFlow's Data API, distribution strategies API, TF Transform and TF-Serving, which leads onto deploying onto Google Cloud AI Platform (or mobile devices)

It seemed to be a great introduction to a large aspect of ML, from organising your data, setting up training and testing data, understanding how to label and characterise your data, before running through an ML model.

The examples given are clear, with terminology that is easy to understand, and includes unsupervised learning techniques such as dimensionality reduction, clustering and anomaly detection. And you will be able to create autonomous learning agents with Reinforced Learning, including using the TF-Agents library.

As this book is designed for use in jupyter you can use Google Colab ML to work through the examples.

Splendid stuff, and a good reference manual to look back on.