eduzhai > Applied Sciences > Engineering >

Dynamic Knapsack Optimization Towards Efficient Multi-Channel Sequential Advertising

  • Save

... pages left unread,continue reading

Document pages: 23 pages

Abstract: In E-commerce, advertising is essential for merchants to reach their targetusers. The typical objective is to maximize the advertiser s cumulative revenueover a period of time under a budget constraint. In real applications, anadvertisement (ad) usually needs to be exposed to the same user multiple timesuntil the user finally contributes revenue (e.g., places an order). However,existing advertising systems mainly focus on the immediate revenue with singlead exposures, ignoring the contribution of each exposure to the finalconversion, thus usually falls into suboptimal solutions. In this paper, weformulate the sequential advertising strategy optimization as a dynamicknapsack problem. We propose a theoretically guaranteed bilevel optimizationframework, which significantly reduces the solution space of the originaloptimization space while ensuring the solution quality. To improve theexploration efficiency of reinforcement learning, we also devise an effectiveaction space reduction approach. Extensive offline and online experiments showthe superior performance of our approaches over state-of-the-art baselines interms of cumulative revenue.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×