Deep reinforcement learning with the confusion-matrix-based dynamic reward function for customer credit scoring.
Yadong WangYanlin JiaYuhang TianJin XiaoPublished in: Expert Syst. Appl. (2022)
Keyphrases
- reward function
- credit scoring
- reinforcement learning
- confusion matrix
- reinforcement learning algorithms
- credit card
- markov decision processes
- logistic regression
- state space
- inverse reinforcement learning
- optimal policy
- transition model
- function approximation
- support vector machine svm
- markov decision process
- temporal difference
- dynamic programming
- information entropy
- learning algorithm
- markov chain
- similarity measure
- active learning
- model free
- risk assessment
- data sets
- text categorization
- generative model