A Methodology for Training Homomorphic Encryption Friendly Neural Networks

Keyphrases

• Accuracy Degradation 14%
• AlexNet Architecture 14%
• Chest X-ray Dataset 14%
• COVID-19 Detection 14%
• CT Dataset 14%
• Deep Neural Network 14%
• Deep Neural Network Inference 14%
• Encrypted Data 14%
• F1 Score 14%
• Friendly Training 14%
• Health Financing 14%
• Homomorphic Encryption 100%
• Homomorphic Encryption Scheme 14%
• Knowledge Distillation 14%
• Lack of Support 14%
• Network Size 14%
• Neural Network 100%
• Operation Type 14%
• Privacy Concerns 14%
• Privacy-preserving 14%
• Quadratic Polynomial 14%
• Rectified Linear Unit (ReLU) 28%
• Regulated Industries 14%
• ReLU Activation 14%
• Retail 14%
• Size Limit 14%
• SqueezeNet 14%
• Structured Methodology 14%
• Trainable Activation Functions 14%
• Training Methods 14%
• Training Model 14%
• X-ray Computed Tomography (X-CT) 14%

Computer Science

• Activation Function 20%
• Deep Neural Network 40%
• Encrypted Data 20%
• Encryption Scheme 20%
• Homomorphic Encryption 100%
• homomorphic-encryption 60%
• Knowledge Distillation 20%
• Network Inference 20%
• Neural Network 100%
• Pre-Trained Model 20%
• Privacy Concern 20%
• Privacy Preserving 20%
• Regulated Industry 20%
• Size Limitation 20%