Game of threads: Enabling asynchronous poisoning attacks

Keyphrases

• Poisoning Attack 100%
• Asynchronous Training 100%
• Model Accuracy 66%
• Adversary 66%
• Thread Scheduling 66%
• Biasing Attacks 66%
• CIFAR-10 66%
• Accuracy Degradation 66%
• Trusted Execution Environment 66%
• Model Prediction 33%
• Training Set 33%
• Validation Set 33%
• Recognition Task 33%
• Updating Model 33%
• Model Bias 33%
• Training Outcomes 33%
• Asynchronous Algorithms 33%
• Image Recognition 33%
• Security Implications 33%
• U-Net 33%
• Intel SGX 33%
• PyTorch 33%
• Machine Learning Training 33%
• Prediction Rate 33%

Computer Science

• Execution Environments 100%
• Machine Learning 50%
• Model Prediction 50%
• Model Accuracy 50%
• Thread Scheduling 50%
• Deep Neural Network 50%
• Security Implication 50%
• Validation Set 50%