Minimum Uncertainty Based Detection of Adversaries in Deep Neural Networks

Mathematics

• Neural Networks 78%
• Uncertainty 70%
• Detector 59%
• Uncertainty Estimation 27%
• Uncertainty Measure 23%
• Scalability 21%
• Low Complexity 20%
• Superlinear Convergence 19%
• Safety 18%
• Training 17%
• Convex Optimization 17%
• Attack 16%
• Small Perturbations 16%
• Bayesian Approach 16%
• Convergence Rate 13%
• Objective function 13%
• Output 12%
• Optimization Problem 11%
• Perturbation 11%
• Unit 11%
• Metric 11%
• Alternatives 10%
• Context 10%
• Performance 9%
• Approximation 8%

Business & Economics

• Neural Networks 100%
• Sampling 52%
• Uncertainty 49%
• Perturbation 37%
• Convex Optimization 23%
• Scalability 20%
• Bayesian Approach 18%
• Objective Function 15%
• Safety 15%
• Inference 13%
• Approximation 13%
• Attack 12%
• Alternatives 7%
• Performance 6%

Engineering & Materials Science

• Deep neural networks 80%
• Uncertainty 52%
• Detectors 39%
• Sampling 35%
• Convex optimization 17%
• Scalability 13%