liveBook · Manning

Sliding window

As explained in Applied Reinforcement Learning

A sliding window is a method that restricts estimation to observations from the most recent fixed or adaptive time interval.

A sliding window keeps estimation focused on outcomes from a recent fixed or adaptive interval. In changing markets, this prevents stale evidence from dominating decisions after customer tastes, campaign performance, competitors, or seasonality shift. section 5.6

A changing campaign

An advertisement’s click-through rate may look stable yesterday but change today. A sliding window emphasizes recent clicks instead of allowing the full historical record to preserve an outdated estimate. section 5.6

Use sliding windows when:

  • Customer preferences, creative performance, competitive campaigns, or seasonal traffic make older outcomes less representative. section 5.6
  • A decision system needs to remain agile rather than cling tightly to historical evidence. section 5.6
  • Pair the approach with continued exploration, because changing contexts can create blind spots. section 5.7

How Applied Reinforcement Learning covers it

How deep it goes
Defines the method and motivation but stops short of specifying window sizes, update formulas, or implementation details.
Examples
Uses changing ad click-through rates, evolving customer tastes, campaign decay, competitor launches, and seasonal traffic.
What it assumes
Assumes familiarity with bandit decisions and stationary payoffs while motivating adaptation to non-stationary markets.
What you can do after
Evaluate when recent outcomes should replace older evidence in a changing bandit environment.
How it connects
Connects non-stationarity with forgetting, discounting, change detection, resets, exploration, and contextual bandits.

Also explained in

Compare how 8 books explain sliding window →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Applied Reinforcement Learning. The section links point to the parts of the book it draws on.