Using reinforcement learning to optimize the acceptance threshold of a credit scoring model.
Mykola HerasymovychKarl MärkaOliver LukasonPublished in: Appl. Soft Comput. (2019)
Keyphrases
- scoring model
- reinforcement learning
- function approximation
- credit scoring
- reinforcement learning algorithms
- threshold selection
- optimal control
- learning problems
- machine learning
- optimal policy
- markov decision processes
- learning process
- state space
- action selection
- risk analysis
- markov decision process
- multi agent
- credit risk
- temporal difference learning
- multi agent reinforcement learning
- neural network
- policy search
- decision making
- reinforcement learning methods
- function approximators
- learning classifier systems
- dynamic programming