Leading With Data· a Bicycle Guide

Partial book profile

Data Science from Scratch First Principles with Python

Joel Grus

This book teaches the fundamental principles of data science by guiding the reader through implementing core algorithms and tools from scratch using Python.

The full book profile is not available yet.

Get the book →

What it’s about

In a world awash with powerful data science libraries that can feel like black boxes, "Data Science from Scratch" offers a unique, first-principles approach to mastering the field. Instead of just learning how to call a library function, this book teaches you what's happening under the hood by having you build the core tools yourself. Starting with a crash course in Python, you'll journey through the essential mathematics and statistics, then dive into implementing fundamental machine learning algorithms like k-Nearest Neighbors, Naive Bayes, regression, decision trees, and even neural networks, all from scratch. This method builds a deep, robust intuition for how data science works, empowering you to move beyond simply using tools to truly understanding them, making you a more effective and insightful data scientist.

The through-line

Who it’s for
An aspiring data scientist, likely a software developer or analyst with some programming and mathematical aptitude. They want to gain a true, deep understanding of data science principles, not just a superficial ability to call library functions, in order to become a confident and competent practitioner.
The problem
The reader is confronted with a vast ecosystem of complex data science libraries and frameworks that act as 'black boxes,' making it difficult to understand how the algorithms actually work. They feel intimidated and may suffer from imposter syndrome, worrying that simply knowing how to use libraries without understanding the underlying principles doesn't make them a 'real' data scientist. They are frustrated by their inability to see inside the black box.
The plan
  1. Begin with a Python crash course to ensure you have the necessary programming fundamentals.
  2. Build a solid foundation by implementing core mathematical and statistical concepts like linear algebra, statistics, probability, and gradient descent.
  3. Implement a wide range of machine learning models from scratch, including regression, decision trees, k-nearest neighbors, Naive Bayes, and neural networks.
  4. Apply your from-scratch toolkit to practical data science applications like recommender systems, network analysis, and natural language processing.
The payoff
You will possess a deep, foundational understanding of the core concepts and algorithms of data science. · You will be able to confidently explain how various models work, what their assumptions are, and how to debug them. · You will transform from a mere user of data science libraries into a practitioner who understands the principles behind them.

See our guide

Related profiles we’ve built

Featured in these guides

Co-cited books