Learning Predictive Analytics with Python

This is a good book. I do not understand why there are bad reviews for it. I would like to thank the author for the good job! Well done! Unfortunately, the author deleted the datasets the book uses from the Google drive.

Book deserves three to four stars max. It is ok and interesting. It is introduces a lot of concepts but shame it doesn't go a little bit more into details especially in the end of the book when talking about clustering and regression. It is one thing to talk about clustering but there is nothing about what to do with it once it is done.there isnt much discussion about regression tree and random forest algorithms which deserve more such as for example what can one do to improve the algos if thstbdont work well or what other algos are available.perhaps simply the book needs to advise on further reading

It tries to tackle relevant topics with predictive modelling, but contains a lot of typos that for someone who has never taken stats before can be misleading. For example in the photo, under the hypothesis test example, Dm should be 21.2 and not 24. May be more suitable for people who already know stats and just need the python aspect, but even then there are bits that need to be read more than once.

The author is writing about using Python, but even the simplest of Python examples are incorrect. For example, the author fails to explain how Python stores and references elements of arrays -- they are zero-based -- then writes examples assuming storage is one-based that result in incorrect results. This is not a single typo. The error is repeated throughout the book. For example, from page 57: "If one wants to select the first 50 rows of the data frame, one can just write: data[1:50]" Experienced Python programmers will know the result will be 49 values, from array locations 1 through 49, but not including the first row, which is row 0. Throughout the book, code is written to be examples, but the result of executing the code is seldom shown. Indeed, if the code had been run and the result shown, the result would immediately illustrate that the code was incorrect. Throughout the book, the grammar is awkward, punctuation confusing, charts inaccurate, programming non-standard, text does not match illustration, context is changed without warning, etc. The sections on predictive analytics and interpretation of results include discussions that are simplistic and discussions that are overly complex. There is very little that will help readers new to the field. I wanted to like this book. I teach machine learning and I was hoping this would be a book I could recommend to my students. Three previously posted reviews give the book five stars. After my experience, I doubt that any of the three read the book. Certainly none tried to run any of the code. One reviewer gave the book one star, but based that score on the author's choice of Python 2 rather than Python 3. While Python 3 is more recent, not all of the support libraries have been converted from Python 2 to Python 3, and many modelers continue to use Python 2. I dislike posting bad reviews. But this book is at very best a rough draft. The author, the Packt editors, the writer of the foreword, and the reviewer of the text, all know this book is not ready for publication and should have sent it back for revision.

Poor English and lots of typos.