We propose a minimax concave penalized multi-armed bandit algorithm under generalized linear model (G-MCP-Bandit) for a decision-maker facing high-dimensional data in an online learning and decision-making process. We demonstrate that the G-MCP-Bandit algorithm asymptotically achieves the...

Motivated by emerging industry practices, this paper studies the effectiveness of a new advance selling strategy in counteracting strategic consumer behavior: the Preorder Contingent Production (PCP) strategy, where the seller's production decision is contingent on an advance selling target. We...