Multi-armed bandit strategies are primarily used in the fields of Artificial Intelligence, Big Data and Smart Data, and Digital Marketing. They help to make decisions when many options need to be tested simultaneously.
Imagine a slot machine with multiple levers („arms“). Each lever dispenses winnings with different frequencies, but you don't initially know which one is the best. Multi-armed bandit strategies are precisely about this: as you try out which lever yields the highest return, you increasingly learn which one is the best choice. Instead of testing all possibilities equally often, you gradually focus on the most promising options.
A practical example: In digital marketing, a website can display two different advertising banners. With a multi-armed bandit strategy, both banners are initially shown equally often. If the system recognises that one of the banners is particularly well-received, it will be shown more frequently. This continuously optimises advertising effectiveness and saves your company time and advertising budget.
In summary: Multi-armed bandit strategies allow companies to discover which alternative works best faster and with data, without unnecessary experiments and inefficiencies.
Deutsch 
English (UK)