O'Reilly logo
live online training icon Live Online training

Building Recommendation Engines in Python

enter image description here

Topic: Data
Max Humber

It’s hard to avoid recommendation engines these days. At companies from YouTube to Netflix to Spotify to Amazon and beyond, recommendations are helping customers find relevant products and businesses sell more products.

Though recommendation engines are super powerful, they’re pretty simple in principle. Or at least so long as the data you’re using is the famous MovieLens dataset. But if you’ve followed a MovieLens tutorial and still struggle to imagine how to implement your own recommendation engine on your own data, this course is for you.

Expert Max Humber shows you how to use industry-leading open source tools to build your own recommendation engines. Join in to gain the skills to help the businesses you work with become as smart as the Netflixs and Amazons of the world.

What you'll learn-and how you can apply it

By the end of this live online course, you’ll understand:

  • How to build a recommendation engine
  • How to use LightFM to build on top of explicit and implicit data
  • How to validate the performance of recommendation engines

And you’ll be able to:

  • Format your data into something that a recommendation engine can understand
  • Serve predictions to users with the help of a Flask app

This training course is for you because...

  • You’re interested in recommendations and how they work.
  • You want to implement recommendation engines in your organization.
  • You want to add recommendation engine experience to your résumé.
  • You’ve tried to wrap your head around MovieLens but don’t know where to go from there.
  • You want to explore industry-leading tools for building recommendation engines.

Prerequisites

  • A working knowledge of Python
  • Familiarity with scikit-learn (useful but not required)

Recommended preparation:

Recommended follow-up:

About your instructor

  • Max Humber is a distinguished faculty member at General Assembly and the author of Personal Finance with Python. Previously, he was the first data scientist at Borrowell and the second data engineer at Wealthsimple.

Schedule

The timeframes are only estimates and may vary according to how the class is progressing

Introduction (5 minutes)

  • Group discussion: Have you ever built a recommendation engine?; Is your interest personal or professional?

Scratch (25 minutes)

  • Presentation: Recommendation basics and theory; just enough math to be dangerous; a landscape tour of the available tools
  • Q&A

Structure (15 minutes)

  • Presentation: Finding and loading data; converting and preparing data for the engine
  • Hands-on exercise: Turn a pandas DataFrame into a sparse matrix

Explicit (20 minutes)

  • Presentation: Building with LightFM; evaluating the performance of the system
  • Hands-on exercise: Build your first model with LightFM

Break (10 minutes)

Implicit (15 minutes)

  • Presentation: Deep learning on implicit feedback; setting up the environment for success
  • Hands-on exercise: Build a system to handle binary responses

Real time (20 minutes)

  • Presentation: Generating recommendations in real time without retraining; serving predictions with Flask
  • Hands-on exercise: Serve recommendations online

Wrap-up and Q&A (10 minutes)