Exploration and Online Learning

CP4285 Modern Recommendation Systems — Week 11 QR code for the Week 11 slide deckScan for the slides
soc-n.us/cp4285-t2610-w11

CP4285 Instruction Team

27 Oct 2026

01 Overview

Week 11: Exploration and Online Learning

Learning outcomes

  • Explain exploration strategies.
  • Design adaptive recommendation policies.
  • Analyze recommendation feedback loops.

  • Discuss ethical implications of experimentation.

02 Adaptive Policies

Multi-armed and contextual bandits

TODO: add exploration-exploitation examples. ## Ethics Thread: Online experimentation ::: {.callout-warning} TODO: examine fair exposure and experimental risk. ::: —

Summary

🔑 Learning online changes what the system chooses to observe.

Next week: LLMs, generative recommendation, and research frontiers.