Skip to content
Datamewby Intura
Start free
Introduction to Machine Learning

Recommendation Systems

Filling the blank cells of a user–item matrix three ways — demographic segments, item features, and similar neighbours — then measuring which one is right.

Level
Intermediate
Lessons
4
Total time
1h 18m
Simulations
5
Included in Introduction to Machine Learning · $9
Tools you'll use
  • Python
  • scikit-learn
  • pandas
  • NumPy
Skills
  • User–item matrix
  • Demographic filtering
  • Content-based filtering
  • Collaborative filtering
  • Matrix factorisation
  • Precision@k and NDCG
Try it now — free

This is the real simulation from the module. No account needed.

What you will be able to do

  • Predict a missing rating with popularity and demographic baselines, and know why both must be beaten
  • Implement content-based and collaborative filtering in ~20 lines of numpy/pandas
  • Explain sparsity, cold start, and why matrix factorisation is the scalable version
  • Evaluate a recommender with rating and ranking metrics, offline and through A/B tests

Before you start

A suggestion, never a lock — you can start here.

  • knn
  • unsupervised-learning
Modules that help first

Syllabus

4 lessons · 1h 18m

Every lesson's title and summary is public. The lessons themselves open once you own the track.

Every lesson in Introduction to Machine Learning

  1. What a Recommender Predicts

    18 min

    Every recommendation is a guess at one blank cell in a user–item matrix — and the easiest guess, the segment average, turns out to be hard to beat.

    Included with the track Simulation

  2. Content-Based Filtering

    18 min

    If you know what an item is made of, you can recommend it to people who liked similar items — even before anyone has rated it.

    Included with the track Simulation

  3. Collaborative Filtering

    22 min

    People who rated alike in the past will rate alike in the future — from that one assumption come user-based, item-based, and matrix factorisation.

    Included with the track Simulation

  4. Evaluating and Combining Recommenders

    20 min

    A smaller RMSE is not necessarily a better list: ranking metrics, A/B tests, hybrid strategies, and the popularity bias that quietly reinforces itself.

    Included with the track Simulation
Lifetime access
$9
Buy the track
More in Introduction to Machine Learning
Module · Beginner

Getting Started

Three kinds of learning, and the two data distinctions that drive almost every later decision.

  • 2 lessons
  • 30 min
  • 2 simulations
Included in Introduction to Machine Learning· $9
Module · Beginner

Dataset Splitting, Cross-Validation, Data Leakage

Before any algorithm. A bad split produces a beautiful number that lies.

  • 2 lessons
  • 34 min
  • 2 simulations
Included in Introduction to Machine Learning· $9
Module · Beginner

Training, Evaluation, Bias–Variance, Tuning

How to train, how to measure with the right metric, and the one trade-off that explains nearly every model failure.

  • 3 lessons
  • 48 min
  • 3 simulations
Included in Introduction to Machine Learning· $9
Module · Beginner

KNN — Classifier and Regressor

The algorithm that learns nothing, and teaches a great deal precisely because of that.

  • 1 lesson
  • 16 min
  • 1 simulation
Included in Introduction to Machine Learning· $9