Python Feature Engineering Cookbook

I was exactly what I needed to know!

As other reviews have stated the book delivers what it says it will; Python code that generates a lot of feature-engineering. I find this book to be fantastic, and Sole's work overall, as it gives life to new feature-engineering possibilities and does it fast. Long gone are the days of writing your own custom transformers or unique time-series features. This book automates a lot of that headache and will absolutely be the first reference I go to when I need to handle a new feature. I personally hadn't dealt with tsfresh prior to reading through and it brought to life instantaneous time-series features I no longer have to write scripts for. A very happy customer on that knowledge alone! Per usual, Sole continues to advance the ML community for the betterment of all.

This book contains all the recipes that are needed for any aspiring data scientist. It contains very good examples that are easy to follow with a good theory explanation on what you are doing.Some basic python knowledge is needed before hand as it wont start from scratch, it is assumed that you have already faced issues with your feature engineering pipelines.The author of this book has created a master piece of art with the feature engineering library, very easy to use and with awesome results.This book became one of my favorite ones very fast!! A must read if you are pursuing a DS/ML/AI position

The Python Feature Engineering Cookbook (PFEC) delivers exactly what the name implies. It’s a collection of recipes targeted at specific tasks; if you’re working in an AI or ML environment and have a need to massage variable data, handle math functions, or normalize data strings, this book will quickly earn a place on your shelf. Each recipe is presented in a standardized format that walks you through the theory and implementation of the code performing the function. Short introductions and appropriate external references provide background for every task, and as long as you have a reasonable familiarity with pandas, scikit-learn, Numpy, Python, and Jupyter, you’ll find a number of uses for the techniques covered.It’s not designed to be a tutorial for those just starting out with machine learning, and isn’t written in a style that invites casual reading. The material tends toward the dry side. While the author does an admirable job of distilling the necessary information into the basic framework of prepare-perform-review, PFEC definitely falls into the reference book category as opposed to being a guide for the uninitiated.In short, you’ll want to have PFEC around if you’re involved in a project that requires hands-on data manipulation in a Python machine-learning environment. Paired with a good guide to ML basics and implementation, it’ll keep you from reinventing quite a few wheels.

Thorough recollection of feature transformations to tackle multiple aspects of data quality and to extract features from different data formats, like text, time series and transactions. Great resource to have at hand when in front of a new dataset.

This book is one of a handful of books covering feature engineering in the market. It covers all the basics and a quick read, so it definitely is an ideal one to get started with. The author also has a video course in Udemy on the same topic which is a good supplement to this book and both can be studied together. The author is very responsive to students' questions in the video course. Once you complete this book, I recommend Max Kuhn's book Feature Engineering and Selection as the next one to deepen your knowledge in feature engineering.

Franchement l eau chaude serait une revolution a cote de ce livre

I bought the kindle verison of book and in introduction the book was good to read but then the display is as in picture , vertically displayed erroneous text . I cannot continue within this impossible display of text . the author need to take care of these issues . Paper verison could be ok i guess i havent read whole book though as i am returning this version now !

Good for getting to know libraries but doesn't explain the statistical context well at all