Efficient and Structural Gradient Compression with Principal Component Analysis for Distributed Training

Computer Science

• Classification Task 50%
• Component Analysis 100%
• Compression Scheme 50%
• Distributed Machine Learning 100%
• Good Performance 50%
• Image Classification 50%
• Machine Learning 50%
• Machine Learning Approach 50%
• Neural Network 50%
• Principal Components 100%
• Speed-up 50%
• Training Data 50%
• Training Process 50%

Chemical Engineering

• Learning System 100%
• Neural Network 25%