Machine Learning and AI Foundations: Recommendations

Introduction

• Welcome
• What you should know before watching this course
• Using the exercise files
• Set up environment

1. The Basics of Making Recommendations

• What is a recommendation system?
• What can you do with recommendation systems?
• Cool uses of recommendation systems

2. Ways of Making Recommendations

• Content-based recommendations: Recommending based on product attributes
• Collaborative filtering: Recommending based on similar users

3. Getting to Know Our Tools

• Introduction to NumPy, SciPy, and pandas
• Think in vectors: How to work with large data sets efficiently

4. Building the Framework for Our Recommendation System

• Explore our product recommendation data set
• Represent product reviews as a matrix
• Recommend by predicting missing user ratings
• A simple way to predict missing user ratings

5. Collaborative Filtering with Matrix Factorization

• Latent representations of users and products
• Code the recommendation system
• How matrix factorization works
• Use latent representations to find similar products

6. Testing Our System

• Explore our system’s recommendations
• Use regularization
• Measure recommendation accuracy

7. Using the Recommendation System in a Real World Program

• Make recommendations for existing users
• How to handle first-time users
• Find similar products

Conclusion

• Wrap up