MENTAL MODEL · M0709

Multi-Armed Bandit

TechnicalHigh supportProbability Theory

Included

Version 1.0.0 · Updated 2026-07-30

CORE DEFINITION

A gambler faces a row of slot machines, each with different and unknown expected rewards. He needs to balance between 'exploration' (pulling new machines to gather information) and 'exploitation' (pulling the machine known to have the highest reward) to maximize total reward.

SCAFFOLDING EFFECT

psychology

Reduce cognitive load

Balance trying and persisting. In early life (or early project stages), adopt a 'high exploration' strategy (more trial and error); after gaining information, gradually shift to a 'high exploitation' strategy (deep cultivation). It provides mathematical guidance for 'when to switch jobs vs. when to go deep'.

anchor

Anchor fast decisions

Multi-Armed Bandit (MAB) trades off between exploration (trying new arms) and exploitation (choosing the current best) to maximize cumulative reward. The mechanism uses probability allocation to balance the unknown and the known.

MINIMUM ACTION

In progress 0/1

Practice this model in one real situation:

Set arms and rewards, use strategies like ε-greedy/UCB to allocate trials, and update online.

Check to track your progress (stored locally)

Learning progress0%

account_treeGenealogyexpand_more

menu_bookReferencesexpand_more

Source support: Explicit

link
en.wikipedia.orghttps://en.wikipedia.org/wiki/Multi-armed_banditZH · Explicit
verified