Worst-case Satisfaction of STL Specifications Using Feedforward Neural Network Controllers: A Lagrange Multipliers Approach

Mathematics

• Feedforward Neural Networks 94%
• Temporal Logic 81%
• Lagrange multipliers 70%
• Controller 65%
• Specification 61%
• Neural Networks 42%
• Reinforcement Learning 34%
• Lagrange multiplier Method 31%
• Percent 27%
• Baseline 24%
• Min-max 24%
• Training 24%
• Maximise 21%
• Robustness 19%
• Initial conditions 18%
• Nonlinear Systems 18%
• Formulation 14%
• Demonstrate 13%
• Framework 13%
• Model 7%

Business & Economics

• Feedforward Neural Networks 100%
• Lagrange multipliers 87%
• Controller 76%
• Neural Networks 54%
• Logic 51%
• Falsification 32%
• Nonlinear Systems 30%
• Reinforcement Learning 28%
• Robustness 28%
• Initial Conditions 26%

Engineering & Materials Science

• Lagrange multipliers 81%
• Temporal logic 80%
• Feedforward neural networks 75%
• Specifications 44%
• Controllers 36%
• Recurrent neural networks 23%
• Reinforcement learning 22%
• Nonlinear systems 19%
• Neural networks 13%