Recently I have been reading a few books from Manning Publications - for those that don't know, Manning produces books with artistic renderings of people on their covers, just like O'Reilly publishes books with animals on top.
Two that caught my eye were these:
First off, when reading any book, the reader has to go in with the right expectations. Neither of these is written with a strong technical focus - on AI or on data science. Rather, they are geared towards managers that need to come up to speed to put in place an optimal structure to enable projects in these areas to be successful. (Not that management is easy by any means, it's just that the skillset needed there is different from that needed to be a competent data scientist.)
What I particularly liked about the first book is that it provides a nice framework to evaluate if your AI-based approach towards solving a business problem is going to a. meet the business needs, and b. deliver sufficient value given the costs that will be incurred. Different books discuss different ways of setting up an AI/ML/DS pipeline, so I am not wedded to a particular model there, but the framework for doing a cost/benefit analysis presented here is quite interesting.
The second book above explains at length how people at different levels in the organization can contribute to the success of a data science project or practice - whether they are leading functions, departments, projects, or tasks.
These are very easy reads without too much technical content. Some might argue that portions of these books would apply to other projects in the workplace as well, not just AI/ML/DS ones, and that would be justified.
One quibble I have regarding these books is the cost. For the value they give you, I think ~$40/book is really quite steep. And I don't think these quite rate as "classics" (not yet, anyway) that I'd want to keep handy on my bookshelf. Perhaps make some notes on the models and frameworks, but that's about it. So yes, these are worth reading - will perhaps take a couple of hours each, but I'd get them from the library. Incidentally, Manning books are (sometimes significantly) cheaper when purchased directly from their webstore (Bing it), than from Amazon.
This is quite technical, but as you dig into it, there is a lot of knowledge, and even some wisdom you can get from it. It starts by giving examples of microservices from some application contexts, and then delves quite deeply into microservice patterns that are quite well explained with helpful diagrams. It also does a very good job describing layering - where you want to solve business problems, where to address the technical issues, how to minimize technical debt, evolve architectures, the relative strengths and weaknesses of various programming and computing paradigms, etc. While not perfect, I'd say it is extremely well done, and highly recommended reading if you need to learn more about microservices.
One book I am looking forward to is this one:
Luis Serrano, the author, has an excellent YouTube channel where he explains complex machine learning algorithms in very simple terms so anyone can understand them. If he carries forward that ability into his book (I haven't read it, but I hope he has), the book should be a pleasure to read.
Speaking of YouTube channels for machine learning, the below two do a good job explaining algorithms and applications. Perhaps not as flamboyant as the Siraj Rawal tutorials, but very good in content nonetheless. See for example RitwickMath's video on Markov Chain Monte Carlo, and Luis Serrano's video on Generative Adversarial Networks. Excellent work!
RitwikMath and Luis Serrano YouTube videos for Machine Learning algorithms
No comments:
Post a Comment