An Excellent Hands-on Introduction to Machine Learning
"A Hands-On Introduction to Machine Learning" by Chirag Shah is a very good data science textbook, starting from the basics, that covers many subjects not usually covered in introductory data science books, including cloud computing, deep learning, dimensionality reduction, bias and fairness for a responsible AI, and a comprehensive coverage of model evaluation (more about it below). The explanations are clear and many insights are present throughout the book.
The book is intended for advanced undergraduates or graduate students, but it can be used by anyone interested in the area, including data scientists wanting to strengthen their skills in a particular area who may want to skip the introductory chapters.
Having a very low barrier for anyone interested in Machine Learning, an entire chapter is dedicated to Python starting from the basics, as the book assumes no prior knowledge or experience with programming. The approach is pragmatic, and some statistics essentials are present in the chapter, which makes for a solid introduction to Python for data science.
The coverage of regression in the fourth chapter is very good, covering very interesting considerations for ML modeling including training time, linearity, number of hyperparameters, and number of features.
Classification is covered in two chapters, interestingly starting with decision tree and random forest and only teaching, among others, logistic regression and SVM in the following chapter. Unsupervised learning receives two chapters covering clustering and dimensionality reduction.
The three chapters on neural network start from basic neural networks and go all the way to convolutional neural network and LSTM and then to using deep models for embeddings, encoders, and transformers. This is an excellent and comprehensive introductory coverage.
An entire chapter is dedicated to designing and evaluating ML systems. It has a compelling section on “Thinking through an ML Solution”. Model evaluation receives a fascinating coverage that includes A/B testing, counterfactual evaluation, and adversarial learning.
The end-of-chapter exercises help a lot in the learning process. They consist not only of questions about the subject matter (conceptual questions), but also of hands-on problems.
Because of its academic nature, having been field tested in university classes, this book is authoritative and clear, with well-thought-out examples and use cases and a coverage that rivals that of the best, more advanced books. I highly recommend this book. After reading it, you will understand why its author, Prof Chirag, has received many awards.
You can buy this great book at Amazon at:
https://amzn.to/4atVUbS
