Abstract
Although a standard reinforcement learning model can capture many aspects of reward-seeking behaviors, it may not be practical for modeling human natural behaviors because of the richness of dynamic environments and limitations in cognitive resources. We propose a modular reinforcement learning model that addresses these factors. Based on this model, a modular inverse reinforcement learning algorithm is developed to estimate both the rewards and discount factors from human behavioral data, which allows predictions of human navigation behaviors in virtual reality with high accuracy across different subjects and with different tasks. Complex human navigation trajectories in novel environments can be reproduced by an artificial agent that is based on the modular model. This model provides a strategy for estimating the subjective value of actions and how they influence sensory-motor decisions in natural behavior.
Highlights
Modeling and predicting visually guided behavior in humans is challenging
The underlying subjective reward value for an action can be estimated through a machine learning technique called inverse reinforcement learning
Standard reinforcement learning methods were developed for artificial intelligence agents, and incur too much computation to be a viable model for real-time human decision making
Summary
Modeling and predicting visually guided behavior in humans is challenging. In various contexts, it is unclear what information is being acquired and how it is being used to control behaviors. In a complex task such as crossing a road, a person must simultaneously determine the direction of heading, avoid tripping over the curb, locate other pedestrians or vehicles, and plan for future trajectory. Each of these particular goals requires some visual evaluation of the state of the world in order to make an appropriate action choice in the moment. A fundamental problem for understanding natural behavior is to be able to predict which subgoals are currently being considered, and how these sequences of visuomotor decisions unfold in time
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have