APPENDIX C

Multi-Armed Bandit (MAB) Simulation Model

We researched the characteristics and effectiveness of each of the approaches to strategy in the strategy palette by simulating how they perform in different business environments. We modeled environments as a so called multi-armed bandit (MAB) problem, which is able to richly capture the economics of decisions under uncertainty. Different algorithmic solutions to this problem then represent the strategic approaches in the palette.

The MAB problem is named after a well-known problem in decision theory. A gambler is faced with choosing which of several slot machines to play. After having played, the gambler will have some knowledge about the payout of some machines but no knowledge of others ...

Get Your Strategy Needs a Strategy now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.