I first came across the notion of the “antilibrary” in Maria Popova’s beautiful post reflecting on “Why Unread Books Are More Valuable to Our Lives than Read Ones“. The term was coined by Nassim Nicholas Taleb (author of The Black Swan) who suggests that as your knowledge grows, so too should your accumulation of unread books. The concept immediately resonated with me as I have always felt purchasing a book bestows the promise of future knowledge. And yet despite this I would also experience mild guilt at seeing that pile of partially read books next to my bed (and more recently the unread thumbnails on my Kindle!). It was both comforting and amusing to learn that the Japanese actually have a term for this phenomenon: “Tsundoku”. Having made peace with not having to read (or even understand) every single page, I’ve become very fond of my library and its accompanying antilibrary – therein lie whole worlds of knowledge just waiting to be explored!

Of course there are many wonderful online resources, but I do still love a good book – whether paper or electronic. In my data science journey these are the ones that have really stood out so far – that I return to over and over again…

Calculus Made Easy: Being a Very-Simplest Introduction to Those Beautiful Methods of Reckoning Which Are Generally Called By the Terrifying Names of the Differential Calculus and the Integral Calculus by Silvanus Phillips Thompson

This book was published in 1914, but as you can tell by the title the author approaches the topic with humour and simplicity.

No bullshit guide to linear algebra by Ivan Savov

I’ve only delved into the small fraction of the linear algebra topics covered – those needed for machine learning applications. But this book makes me want to know more: each topic is explained in such a beautiful, intuitive way.

Naked Statistics: Stripping the Dread from the Data by Charles Wheelan

I read this one cover-to-cover quite quickly: it was a wonderfully non-threatening and fascinating intro to common concepts and pitfalls.

Head First Statistics: A Brain-Friendly Guide by Dawn Griffiths

This was my go-to reference during the initial weeks of my MSc statistics course – it’s a highly visual, practical, and non-threatening book for which I am grateful.

Grokking Deep Learning by Andrew W. Trask

This was the first book on deep learning that actually made sense to me! You work your way up from the math behind a single neuron, a single forward pass, a single iteration of back propagation, through to building your own simple neural network using just Numpy so that the mechanics of the process really start to make sense!

Deep Learning with Python by François Chollet

This was the second book on deep learning that made sense to me! Chollet builds your understanding of neural networks brick by brick, using the ‘universal workflow of machine learning’ as its foundation. The world of Keras and Tensorflow was rather intimidating to me at first but this book ensures you develop a solid foundation of how to develop neural networks, exploring the search space for what works and what doesn’t, as well as understanding different foundational architectures like CNNs, RNNs, GANs, LSTMs, and so on.

Artificial Intelligence: A Modern Approach by Stuart Russell & Peter Norvig

AI is a big topic, and a complex one. But this is the book that breaks it down into its component parts so that a bigger picture starts to emerge through an improved understanding of the individual puzzle pieces.

Linguistic Fundamentals for Natural Language Processing II: 100 Essentials from Semantics and Pragmatics by Emily M. Bender & Alex Lascarides

As an aspiring NLP practitioner with no linguistics background this book fills a gap! Short and often thought-provoking snippets provide a window into the many concepts that linguistics defines and grapples with.

Speech and Language Processing by Dan Jurafsky & James H. Martin

As far as I’m concerned this is the definitive textbook for all things NLP, covering a range of concepts from regex to LLM’s in great detail and with intuitive explanations that I found very helpful. Complement with the Stanford CS224N playlist on YouTube.