Collaborative Filtering and Embeddings — Part 1

Shikhar Gupta
Towards Data Science
7 min readDec 28, 2017

Recommendation systems are all around us. From Netflix to Amazon to even Medium, everyone is trying to understand our taste so that they can drive us into continuous engagement. You’ll be amazed by the amount of work that goes behind this. Let’s try to understand the mechanics of one such way of developing recommendation systems.

Introduction

In this series of articles I’ll explain collaborative filtering, a very common technique for developing automated recommendation systems (although the use case is not limited to recommendations). However, what makes this discussion interesting is how we can understand a more general concept of embeddings while implementing this technique. If you haven’t heard about this term no need to worry, you’ll have a good idea by the end of this post.

Part 1:

  • Basic idea behind collaborative filtering
  • A simple algorithm to implement collaborative filtering in excel (yeah you read it right!!)
  • Understanding the concept of embeddings and bias

Part 2:

  • Implementation of collaborative filtering using fastai library
  • Interpreting and visualising embeddings

--

--

Towards Data Science
Towards Data Science

Published in Towards Data Science

Your home for data science and AI. The world’s leading publication for data science, data analytics, data engineering, machine learning, and artificial intelligence professionals.

Shikhar Gupta
Shikhar Gupta

Written by Shikhar Gupta

Applied scientist @Amazon, USF MSDS & IIT Roorkee alumunus (Twitter:@shik1470)

Responses (7)

What are your thoughts?