Books

I’d like to highlight a couple technical leaning books I’ve found either helpful or illuminating that I don’t think get mentioned that much. There are several other more popular books I really enjoy (e.g. Fooled by Randomness, Introduction to Statistical Learning, The Visual Display of Quantitative Information), but I often don’t see the following talked about.

Debugging: The 9 Indispensable Rules for Finding Even the Most Elusive Software and Hardware Problems

Early in my software career I found that I spent a lot of my time debugging code, but that wasn’t something I learned much about in school. I searched for resources on specific techniques to better help with this task, and fortunately wound up reading Debugging. The structured approach it describes took what was often a frustrating task to one of my greatest strengths.

The steps here also don’t only apply to software development. Any system that is not working as expected from your TV at home, a data report at work, etc. can utilize the techniques here to isolate and find what is causing your errors.

Observation and Experiment

This is the best written book of a technical topic I have ever read. The language is incredibly clear and understandable while still being completely formal, robust, and precise. The subject of the book is causal inference and as the title suggests it is divided into two parts: Randomized Experiments (Experiments) and Observational Studies (Observation). The first part on experiments is a mere 60 pages but covers the topic so well it was usually the only reference I needed when designing A/B tests at Clio. Observational studies is unfortunately much more difficult topic, as illustrated by its part being about 4 times longer than the experiments section :D

The Art of Scientific Investigation

The paradoxical title alludes to the fact that scientific investigation is often not obvious or prescriptive and has numerous struggles, road-blocks, and challenges that requires ingenuity and creativity to overcome. Written in the 1960s with many anecdotes from decades before, it’s shocking how many parallels there are between early scientific discover and the modern business environment. Each chapter brings new insights and strategies that will help further your own (scientific) inquiry.

Introduction to Probability

The book All of Statistics starts with this figure:

Most inference and data mining is trying to “reverse” the data generating process, which is useful to assume stemmed from a probabilistic model. So if you don’t know probability, this is going to be confusing and hard :D

The textbook itself is excellent, but one of the authors also teaches an associated course at MIT, which is available online on edX. The course is tough and requires many hours of practice beyond the graded exercises, but passing will ensure an excellent grasp of probability that is indispensable and will be useful in most further studies of Statistics. edX also uses a type of symbolic processing for answering quiz/test questions, which offers the ability of asynchronous evaluation while still requiring students to do all the calculations and come up with the answers themselves (unlike multiple choice questions).

If anyone is looking to expand beyond this, the subsequent statistics courses is also of similar (high) quality. If you’re interested in Statistics or Data Science these courses are worth many times their cost.