Impression Allocation for Combating Fraud in E-commerce Via Deep Reinforcement Learning with Action Norm Penalty.
Mengchen ZhaoZhao LiBo AnHaifeng LuYifan YangChen ChuPublished in: IJCAI (2018)
Keyphrases
- reinforcement learning
- action selection
- penalty term
- objective function
- partially observable domains
- online banking
- electronic commerce
- reward shaping
- action space
- function approximation
- resource allocation
- state space
- optimal allocation
- markov decision processes
- agent learns
- transition model
- fraud detection
- state action
- data mining techniques
- model free
- reinforcement learning algorithms
- anomaly detection
- multi agent
- machine learning
- temporal difference
- dynamic programming
- policy search
- data mining
- partially observable
- combinatorial auctions
- human actions
- internet banking
- learning algorithm