To calculate the overall star rating and percentage breakdown by star, we don’t use a simple average. Instead, our system considers things like how recent a review is and if the reviewer bought the item on Amazon. It also analyses reviews to verify trustworthiness.
I got book today. Surprised to see the quality of the book. No color picture and pages look like photocopy with poor quality ink. Quite disappointed as not getting motivation to start reading. Be careful before you order.
In the jungle of publications about ML, this book provides a unique hands-on and principled set of tools to really get you through a project from start to finish. A must read to any working data scientist or data engineer out there. Can't recommend it enough.
I don't think the author has built a machine-learning powered application. This book is extremely lightweight at a little over 200 pages and is too high-level to have any practicality. The content is just an odd assortment of stuff with bizarre sidebars on transfer learning and code snippets with no cohesiveness. The chapter on deployment is exactly ten pages long and is a big nothing burger. I don't even recommend this book for a beginner because it will confuse them.
I will start off by saying on a scale of 1 to 10 in data science / machine learning knowledge (1 being "I barely know what a linear model" is and 10 being "I contribute to building Machine Learning Libraries / conduct research") that I am around a 4. I initially bought this book because I have a decent understanding of Data Science (created a few models at work and personally) and was interested in ways to serve the model via webserver like flask/django.
The best analogy I can give about this book is its like going to a restaurant seeing "beef stew" on the menu and ordering it. When it arrives, you realize it is just beef broth and when you complain to the waiter they tell you "beef was stewed in it, but you have to pay extra for the actual beef". Hence the title of my review.
Chapter after chapter I kept waiting for him to dive into the python scripts and explaining how they build the model. In this 250 page book, maybe 30 of the pages are dedicated to explaining the model and pipeline with the rest dedicated to superficially explaining DS/ML concepts.
It doesn't go deep enough for anyone who has an intermediate level of knowledge DS/ML. On the other hand, it doesn't explain enough for people who might be beginners. For example, it just assumes you understand when to apply XGBoost versus using Scikit-Learn. But then on the next page it tries to explain the K Nearest Neighbors algorithm. Like you are expecting the reader to understand how different Machine Learning libraries affect computational needs, but then assume they don't know the most basic clustering algorithm? What???
To me it feels more like a hastily written, half white paper/half wiki article about DS/ML algorithms, Computer Science, and how Machine Learning is actually bad for humanity.
Also, he interviews people who have DS/ML experience, which is a good idea and cool in theory, but some of the interviews just feel like sales pitches for their products. Like, I haven't used StictchFix, and it might be a great product, but I will go to their website to learn about it. I don't want to pay to read a sales pitch.
I wish I could return this book, but have already highlighted it from from to back. Please don't buy this book unless you fall into whatever very niche group this author targeted the book towards. Instead buy Hands on Machine Learning if you want to learn more about DS/ML. If you want to know how to deploy your models, maybe try Applied Data Science 2.0, but due to version updates and dependencies I couldn't get it to deploy, but the reference on how to build the pipeline is useful.
To me this book felt like a lot of bad Medium or Towards Data Science articles stacked on top of each other.
This book is NOT an overly technical book. The way I read it, it's a book that's centered around the lessons the author, Emmanuel , learned during his time as a data scientist/ML engineer. He formats these lessons in such a way that makes the book extremely easy to read and grasp. As a newly-hired data scientist who has been charged with created the company's anomaly detection application, this book will serve me well!
I've met a lot of people who would say they are well aware of the contents of this book and that they would have nothing to learn from reading it. But, it amazes me how many times I've seen those people spin up projects and completely ignore the steps they claim to know. If you're managing a team, I think this should be required reading. "Building Machine Learning Powered Applications: Going from Idea to Product" helps to crystalize the best practices that are, all too often, neglected at fast-moving startups and on rapid-prototyping teams.