Best books for Data Scientists

Python for Data Analysis, 3rd Edition

by Wes McKinney · 2011

~550 pages~14 hrs to read

Written by the creator of pandas, it covers loading, cleaning, reshaping and summarising tabular data in Python. Read it early, alongside real datasets, because nearly every later data task starts with these skills.

Practical Statistics for Data Scientists, 2nd Edition

by Peter Bruce, Andrew Bruce, Peter Gedeck · 2017

~370 pages~9 hrs to read

Explains the statistics a working data scientist needs: sampling, distributions, A/B testing, regression and classification, with code in R and Python. Read it when you can already wrangle data and want to know which method fits the question.

An Introduction to Statistical Learning: with Applications in Python

by Gareth James, Daniela Witten, Trevor Hastie, Robert Tibshirani, Jonathan Taylor · 2013

~620 pages~16 hrs to read

A clear first course in regression, classification, resampling, trees and other core statistical learning methods, with Python labs. It is the best place to build the intuition behind the models you will fit every day.

Build a Career in Data Science

by Emily Robinson, Jacqueline Nolis · 2020

~250 pages~6 hrs to read

Covers the job itself: finding a first role, interviewing, working with stakeholders, managing projects and moving up. Read it in your first months, or while job hunting, to learn what the work looks like beyond the modelling.

Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow, 3rd Edition

by Aurélien Géron · 2017

~860 pages~22 hrs to read

Project-driven walk-through of the machine learning workflow, from classical models in Scikit-Learn to neural networks in Keras. Work through it once you are comfortable with Python and basic statistics.

Feature Engineering for Machine Learning

by Alice Zheng, Amanda Casari · 2016

~220 pages~6 hrs to read

A compact guide to turning raw numeric, text and categorical data into features that models can use. Useful once your first models plateau and the gains have to come from the data rather than the algorithm.

Storytelling with Data: A Data Visualization Guide for Business Professionals

by Cole Nussbaumer Knaflic · 2015

~290 pages~7 hrs to read

Teaches how to choose the right chart, remove clutter and build a narrative around a result. Read it before your next stakeholder readout, since an unclear chart is the usual reason a good analysis gets ignored.

The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition

by Trevor Hastie, Robert Tibshirani, Jerome Friedman · 2001

~750 pages~19 hrs to read

The rigorous, mathematical companion to the introductory text, covering regularisation, boosting, additive models and more. Keep it as a reference for when you need to understand why a method behaves as it does.

Data Science for Business: What You Need to Know about Data Mining and Data-Analytic Thinking

by Foster Provost, Tom Fawcett · 2013

~410 pages~10 hrs to read

Frames data science as a way to make business decisions, covering problem framing, evaluation and the trade-offs behind predictive models. A good fit for senior data scientists who must scope projects and explain their value to non-technical leaders.

Designing Machine Learning Systems: An Iterative Process for Production-Ready Applications

by Chip Huyen · 2022

~390 pages~10 hrs to read

Covers data pipelines, deployment, monitoring and the lifecycle of production models. Read it when your work shifts from notebooks to models that other teams depend on.

Trustworthy Online Controlled Experiments: A Practical Guide to A/B Testing

by Ron Kohavi, Diane Tang, Ya Xu · 2020

~200 pages~5 hrs to read

The reference on running A/B tests that you can believe: metric choice, pitfalls, sample ratio checks and experiment platforms. Read it when you own experimentation or advise a team on it.

Causal Inference: The Mixtape

by Scott Cunningham · 2021

~580 pages~15 hrs to read

Introduces difference-in-differences, regression discontinuity, instrumental variables and related methods for estimating causal effects from observational data. Read it when stakeholders ask what an intervention actually caused, not just what correlates with it.