The folks working in the field of Information Technology and related industries would be well aware of 'Machine Learning'. Hardly a day goes by without us reading one or the other news about how the machines are getting better at analyzing, predicting, and in some cases taking over the various fields once dominated by us.
If like me, you too are a newbie to this Next Big Thing, and wondering about where you can get a bird's eye view, this book is a good place to start.
In 'The Master Algorithm', author Pedro Domingos begins with real-life, familiar examples of how ML pervades our lives even today. Next, we get an overview of various high-level methodologies that comprise ML. The book concludes with an insight into a 'master algorithm', a mind-boggling concept, one that means the rise of learning machines that would create more learning machines.
This book makes for a persuasive and captivating read. Of course, it barely scratches the tip of a giant iceberg, and the author is quick to admit that the current techniques may not only evolve vastly but in many cases, there could be completely unforeseen breakthroughs.
No matter what your area of work is, one thing seems certain: Machine Learning is going to impact all of us, it is here to stay and evolve into general artificial intelligence.
As humans, we better be scared... And be better prepared.
Garry Kasparov is a multifaceted personality. He is an author of several critically-acclaimed books, much sought-after orator, political opponent to none other than the mighty Vlad Putin... And, did I mention that he is a former World Chess Champion and widely regarded as one of the greatest chess players of all time?
I have been a die-hard Kasparov fan (except when he was playing against the Indian chess prodigy V. Anand), and yet I was reluctant to pick up this book, thinking it to be some half-baked pseudo-science.
Boy, was I so happy to be proved thoroughly wrong!
"Deep Thinking" has three main threads running in parallel throughout the book.
At the core of it, Kasparov writes about his chess 'wars' with Deep Blue, the IBM supercomputer.
On top of this, he elaborates the history of chess-playing machines: Deep Junior, Hydra, Fritz, et al.
Encompassing this, he muses about machine learning, human-machine interfaces and artificial intelligence in general.
You need not be a chess pro or an AI expert to enjoy this book. Kasparov has an engaging writing style, and his razor-sharp, no-holds-barred approach makes for a delightful read!
Do you like algorithms? And, like me, do you often get confused and frustrated with how to deal with humans in a rational, logical way? And.. Do you often wish there were computer algorithms to deal with human behaviour and interactions?
If your answer to these questions is No, then don't read this book.
"Algorithms to live by: The computer science of human decisions" attempts to reduce complex, often chaotic human actions and decision-making processes to a set of well-defined mathematical rules.
For example, how many houses you should look at before deciding which one to buy? (Short answer: 37%). Or which is the best way to sort your books? Or to schedule interviews of candidates so that you get the best person in shortest time? Or how to use the concept of caching to organize your papers in an optimum manner?
The book is full of such examples of how to use the mathematical theories that are commonly used in computer algorithms, into our day to day lives. The authors, Brian Christian and Tom Griffiths, have taken care not to faze and daze the reader with too much jargon, which makes this book an easy read.
The only complaint I have about this one is: Even after reading 370 pages, I wanted more of it. That, perhaps, is less of a complaint and more of an acknowledgement of how good this book is.
