11 Result(s)
-
Article
Global optimization of objective functions represented by ReLU networks
Neural networks can learn complex, non-convex functions, and it is challenging to guarantee their correct behavior in safety-critical contexts. Many approaches exist to find failures in networks (e.g., adversa...
-
Article
Guest Editorial: Special issue on robust machine learning
-
Article
Generating probabilistic safety guarantees for neural network controllers
Neural networks serve as effective controllers in a variety of complex settings due to their ability to represent expressive policies. The complex nature of neural networks, however, makes their output difficu...
-
Chapter and Conference Paper
ZoPE: A Fast Optimizer for ReLU Networks with Low-Dimensional Inputs
Deep neural networks often lack the safety and robustness guarantees needed to be deployed in safety critical systems. Formal verification techniques can be used to prove input-output safety properties of netw...
-
Chapter and Conference Paper
Normalizing Flow Policies for Multi-agent Systems
Stochastic policy gradient methods using neural representations have had considerable success in single-agent domains with continuous action spaces. These methods typically use networks that output the paramet...
-
Article
Decomposition methods with deep corrections for reinforcement learning
Decomposition methods have been proposed to approximate solutions to large sequential decision making problems. In contexts where an agent interacts with multiple entities, utility decomposition can be used to...
-
Chapter and Conference Paper
The Marabou Framework for Verification and Analysis of Deep Neural Networks
Deep neural networks are revolutionizing the way complex systems are designed. Consequently, there is a pressing need for tools and techniques for network analysis and certification. To help in addressing that...
-
Chapter and Conference Paper
Robust Super-Level Set Estimation Using Gaussian Processes
This paper focuses on the problem of determining as large a region as possible where a function exceeds a given threshold with high probability. We assume that we only have access to a noise-corrupted version ...
-
Chapter and Conference Paper
Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks
Deep neural networks have emerged as a widely used and effective means for tackling complex, real-world problems. However, a major obstacle in applying them to safety-critical systems is the great difficulty i...
-
Chapter and Conference Paper
Collision Avoidance Using Partially Controlled Markov Decision Processes
Optimal collision avoidance in stochastic environments requires accounting for the likelihood and costs of future sequences of outcomes in response to different sequences of actions. Prior work has investigate...
-
Chapter and Conference Paper
Evolving Hierarchical and Recursive Teleo-reactive Programs through Genetic Programming
Teleo-reactive programs and the triple tower architecture have been proposed as a framework for linking perception and action in agents. The triple tower architecture continually updates the agent’s knowledge ...
