International business machines corporation (20240211746). REALISTIC SAFETY VERIFICATION FOR DEEP REINFORCEMENT LEARNING simplified abstract
Contents
REALISTIC SAFETY VERIFICATION FOR DEEP REINFORCEMENT LEARNING
Organization Name
international business machines corporation
Inventor(s)
Kevin Eykholt of White Plains NY (US)
Wenbo Guo of State College PA (US)
Taesung Lee of Ridgefield CT (US)
Jiyong Jang of Chappaqua NY (US)
REALISTIC SAFETY VERIFICATION FOR DEEP REINFORCEMENT LEARNING - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240211746 titled 'REALISTIC SAFETY VERIFICATION FOR DEEP REINFORCEMENT LEARNING
Abstract: Safety verification for reinforcement learning can involve receiving a policy generated by deep reinforced learning, which is then used to act in an environment with a set of states. If the policy is non-deterministic, it can be decomposed into a set of deterministic policies. If the state-transition function associated with the states is unknown, it can be approximated by training a deep neural network and transforming it into a polynomial. The policy with the state-transition function can then be verified using a constraint solver, and runtime shielding can be implemented.
Key Features and Innovation:
- Safety verification for reinforcement learning
- Decomposition of non-deterministic policies into deterministic policies
- Approximation of unknown state-transition functions using deep neural networks
- Verification of policies using constraint solvers
- Implementation of runtime shielding
Potential Applications: This technology can be applied in various industries such as autonomous vehicles, robotics, gaming, and industrial automation to ensure safe and reliable decision-making processes.
Problems Solved: This technology addresses the challenge of verifying the safety of policies generated by reinforcement learning algorithms in complex and uncertain environments.
Benefits:
- Enhanced safety and reliability in decision-making processes
- Improved performance of reinforcement learning algorithms
- Increased trust in autonomous systems
Commercial Applications: The technology can be utilized by companies developing autonomous systems, robotics, and AI-powered applications to ensure safety and reliability in their products, leading to increased market competitiveness and customer trust.
Questions about Safety Verification for Reinforcement Learning: 1. How does the technology verify the safety of policies in reinforcement learning? 2. What are the potential implications of using deep neural networks to approximate unknown state-transition functions in reinforcement learning algorithms?
Frequently Updated Research: Researchers are continuously exploring new methods and techniques to improve the safety and reliability of reinforcement learning algorithms, including advancements in policy verification and approximation of state-transition functions.
Original Abstract Submitted
safety verification for reinforcement learning can include receiving a policy generated by deep reinforced learning, where the policy is used in acting in an environment having a set of states. responsive to determining that the policy is a non-deterministic policy, the non-deterministic policy can be decomposed into a set of deterministic policies. responsive to determining that a state-transition function associated with the set of states is unknown, the state-transition function can be approximated at least by training a deep neural network and transforming the deep neural network into a polynomial. using a constraint solver the policy with the state-transition function can be verified. runtime shielding can be performed.
