Browse by topic
01Foundations
Frame an ML problem
Define the decision before choosing a model.
Features and training data
Prevent the model from learning information it will not have at serving time.
Embeddings and retrieval
Represent similarity with a learned vector space.
Evaluation and generalization
Measure the mistakes that matter.
02Production designs
Design video recommendations
Separate candidate generation from ranking.
Design content moderation
Combine automation with explicit escalation paths.
Design bot detection
Reason about adversarial adaptation.