Do i need to take a TinyML course in edx after reading TinyML book by Pete Warden and Daniel Situnayake?
Or is it enough just to read a book
it’s free so have a look through it at least. only pay if you want a certificate
I like to think of the book as the textbook for the edX course series – some people only need to read the textbook and never attend class – others learn more from class – except in this case the correspondance isn’t exactly 1 to 1 so if you skim through the edX course I’m sure there will be at least a few sections with new interesting information.
Also we just released course 4: MLOps for Scaling TinyML which is entirely new content beyond the book! https://www.edx.org/course/mlops-for-scaling-tinyml
I’m happy to share my experiences, so your mileage may vary.
I studied software engineering and maths at undergraduate uni and am mainly into arduino programming as an expensive hobby. I got into TinyML from the book and got a few of their examples working, but if I had to be honest I didn’t quite understand why I was doing in the Colab environments, I was just “following the bouncing ball”. I also had a similar experience using Edge Impulse… which is awesome btw.
As I didn’t have an ML background outside of some statistics I did at uni, I felt the course gave me a great grounding in the fundamentals of machine learning in a familiar context (ie embedded software). And understand some of the non technical aspects of machine learning like ethics and diversity.
So my feelings is that if you are experienced in both machine learning and arduino programming the book is more than enough, but if you don’t have experience in either ML or arduino programming then the course is definitely worth the time and energy investment.
As a side note, I found that the course helped me critically analyse other machine learning projects. But as a disclosure note I did complete the professional TinyML certificate and have a copy of the TinyML book. Hope it helps
Glad you found the course helpful and useful! I hope you now get to make some TinyML powered “expensive hobby projects” lol