Worth Your Time: "Machine Learning For Robotics"

First, the link to O’Reilly to acquire the book “Machine Learning For Robotics” by Alishba Imran and Keerthana Gopalakrishnan is here.

Now that we have that out of the way, here’s

Three reasons why “Machine Learning For Robotics” Is Worth Your Time:

1. This book is a new, solid foundation for the intersection of Machine Learning and Robotics

   There are plenty of books (and Twitter “influencers”) that talk about either machine learning or robotics. However, despite the large amount of overlap these fields have, they’re still distinct. This is primarily because machine learning is rooted in statistics, and robotics is rooted in control theory/dynamics. (Both are rooted in maths, so if that isn’t a sign from the universe then I don’t know what is).

   Due to these differences, there aren’t many resources that talk about how machine learning is affecting robotics and vice versa. Therefore, since many folks come into this intersection from either the perspective of being an ML researcher or a roboticist, they often don’t have the strongest base.

   I think this book will help alleviate that problem.

2. This book has a good sense of history.

   “Wow, look at this new paper, it’s so old.”

   “When was it published?”

   “Last week.”

   Such is the reality of modern AI research: move fast and break things, but to build much better things it’s important to understand what’s come before.

   This book does a fantastic job of exploring where our modern algorithms actually come from, and I’m especially a fan of how they introduced transformers: through the history of recurrent neural networks and their ilk.

3. This book interconnects topics well.

   This is somewhat related to my first point in that if you’re doing an intersectional textbook, then you’re going to have to cover a lot of ground. Making a good foundation is hard work, and the difficulty lies in knowing how to make sure your reader knows the million things that make up both fields without using a million pages.

   I never felt as though concepts came “out of nowhere” reading this book. There were always plenty of definitions, demonstrations, and examples: grounding wires to make sure that I could navigate the storm that is this exciting new field of intelligent robotics.

This book isn’t even planned to be finished until 2024, so there will be many more chapters and things to learn. However, if the following chapters are as good as the first two, then I’ll be happy to keep on reading.

If you enjoyed this post, then later this week you’ll enjoy my in-depth review of Eric Jang’s new book “AI Is Good For You,” a book that’s also about AI and Robotics, but from Eric’s perspective as a world-class scientist and startup executive. You can learn more about his book here and see my initial review here.