Offline Deep Reinforcement Learning for Dynamic Pricing of Consumer Credit
This addresses the problem of personalized pricing in consumer credit for financial institutions, offering a method that avoids costly online experiments, but it is incremental as it applies existing offline RL techniques to a specific domain.
The paper tackled dynamic pricing of consumer credit by applying offline deep reinforcement learning to a static dataset, eliminating the need for online interaction or price experimentation, and demonstrated that their method using conservative Q-Learning learned an effective personalized pricing policy on real and synthetic data.
We introduce a method for pricing consumer credit using recent advances in offline deep reinforcement learning. This approach relies on a static dataset and requires no assumptions on the functional form of demand. Using both real and synthetic data on consumer credit applications, we demonstrate that our approach using the conservative Q-Learning algorithm is capable of learning an effective personalized pricing policy without any online interaction or price experimentation.