Data science is an emerging discipline spurred by the need to extract meaningful information and gain insights from large amounts of complex data. Data science encompasses many areas from mathematics, statistics, and information and computer sciences.

This book contains a broad range of timely topics and presents interesting examples on real-life data using Python. Python is an interpreted, dynamically typed programming language widely used by data scientists. Python has libraries for scientific computing and numerical analysis that support array manipulation and plotting, data frame and associated manipulations, as well as modules for machine learning and data mining.

The book has 11 chapters. Two chapters give a brief introduction and a primer on Python. Six chapters present concepts and techniques on descriptive statistics, statistical inference, regression analysis, machine learning (supervised, unsupervised), and parallel computing. Three chapters describe applications of recommender systems, network analysis, and sentiment analysis.

Every chapter except for the introduction has Python code that can be downloaded from the GitHub [1] server in the form of Jupyter notebooks [2], which include live code, equations, visualizations, and narrative text that enable hands-on experimentation.

According to the authors, the goal of the book is “to offer a panoramic view of the data science field.“ This may sound fair with respect to the coverage of concepts and techniques, but it is a stretch when it comes to the coverage of applications. The authors acknowledge leaving out advanced concepts and techniques such as big data analytics, deep learning, and advanced mathematics and statistical methods, but do not say anything about the lack of coverage of the various application categories or about how applications in the book fit into the big picture.

The presentation is rather terse and glosses over the background information needed to make sense of the solutions implemented in the code. The emphasis is more on Python and on how to implement solutions to data science problems rather than the principles and the knowledge needed to come up with the solutions in the first place. The authors only give references so that the reader may “delve deeper” into the topics in each chapter.

The book could be improved if the authors include exercises at the end of each chapter and outline the process and the effective extent to which experience gained from use cases in the book affect learning and performance in a new situation.

The book is loosely organized with weak threads binding the various chapters. The authors give little rationale for the order of the chapters and the selection of their content. They give no explanation of the merit of Python and its pros and cons versus, say, the R statistical computing framework [3].

Overall, the book is a good addition to references on Python and data science. Students as well as practicing data scientists and engineers will benefit from the many techniques and use cases presented in the book.

More reviews about this item: Amazon