Multi-armed Bandit Problem with Known Trend
This incremental work addresses online decision-making problems like active learning and recommendation systems, offering specific gains for scenarios with predictable reward trends.
The paper tackles the multi-armed bandit problem with known trend, where reward functions have known shapes but unknown distributions, by proposing the A-UCB algorithm, which shows improved regret bounds and experimental performance compared to UCB1.
We consider a variant of the multi-armed bandit model, which we call multi-armed bandit problem with known trend, where the gambler knows the shape of the reward function of each arm but not its distribution. This new problem is motivated by different online problems like active learning, music and interface recommendation applications, where when an arm is sampled by the model the received reward change according to a known trend. By adapting the standard multi-armed bandit algorithm UCB1 to take advantage of this setting, we propose the new algorithm named A-UCB that assumes a stochastic model. We provide upper bounds of the regret which compare favourably with the ones of UCB1. We also confirm that experimentally with different simulations