In the tech world; I believe sometimes you try to fly before you can run! (No puns intended).
Fresh from my Python learning, the first thing on my mind was how to elope into the world of Machine learning, it was like I was liberated from prison and my first hunger to fill was the appetites of Deep Learning!
Of course, I wasn’t a novice on the subject matter: for it was the curiosity of AI that led me to begin to code in the first place. I had heard my fair share of podcasts; starting with Lex Fridman’s(a research scientist at MIT): – “Artificial Intelligence”: “Talking Machines” hosted by Katherine Gorman(nerd, journalist) and Ryan Adams(nerd, Harvard CS professor). In short, I had my bearings well informed before I picked up this read.
“Fundamentals of Deep Learning,” is the most eloquently written tech book I have experienced thus far! From the inception of neural network chapter to the deep reinforcement learning section; I never found the writing dull. The points were short, potent & concise. There was no beating around the bush much: – which favored my stlye of reading. The grammar and style of writing was amazing; saying so knowing full well: – that I am an author myself. # There was a great hint of intelligence on the literary standards of the word.
I don’t think this book was meant to be a beginner’s guide, leave alone the interesting style of writing that most could find hard, this depends on whom you ask though!
The mathematical annotations were well illustrated, so were the flow chart diagrams. The caliber of art on those two things aforementioned were beautiful, that’s something to say on other technical books depictions on the same subject matter.
The models depend a lot on mathematics: and a lack of knowledge on that discipline, hampers a lot on internationalization of the intuition to be gained from this book.
To aggregate the weights of this review; while reading the concepts – I had a lot of inspirations of the applicability of the models on business ideas I already had, they fit like a glove, models to ideas! To that, I see a grand union that will keep iterating of course as par new models being developed. UNREAL(UNsupervised REinforcement and Auxiliary Learning) was a favorite in the book.