Leveraging Machine Learning to Automatically Derive Robust Planning Strategies from Biased Models of the Environment
- Anirudha Kemtur, Rationality Enhancement Group, Max Planck Institute for Intelligent Systems, Tübingen, Germany
- Yash Raj Jain, Rationality Enhancement Group, Max Planck Institute for Intelligent Systems, Tübingen, Germany
- Aashay Mehta, Rationality Enhancement Group, Max Planck Institute for Intelligent Systems, Tübingen, Germany
- Frederick Callaway, Princeton University, Princeton, New Jersey, United States
- Saksham Consul, Rationality Enhancement Group, Max Planck Institute for Intelligent Systems, Tübingen, Germany
- Jugoslav Stojcheski, Rationality Enhancement Group, Max Planck Institute for Intelligent Systems, Tübingen, Germany
- Falk Lieder, Rationality Enhancement Group, Max Planck Institute for Intelligent Systems, Tübingen, Germany
AbstractTeaching clever heuristics is a promising approach to improve decision-making. We can leverage machine learning to discover clever strategies automatically. Current methods require an accurate model of the decision problems people face in real-life. But most models are misspecified because of limited information and cognitive biases. To address this problem we develop strategy discovery methods that are robust to model misspecification. Robustness is achieved by modeling model-misspecification and handling uncertainty about the real-world according to Bayesian inference. We translate our methods into an intelligent tutor that automatically discovers and teaches robust planning strategies. Our robust cognitive tutor significantly improved human decision-making when the model was so biased that conventional cognitive tutors are no longer effective. These findings highlight that our robust strategy discovery methods are a significant step towards leveraging artificial intelligence to improve human decision-making in the real world.