Data-Independent Structured Pruning of Neural Networks via Coresets

Keyphrases

• Neural Network 100%
• Data Independence 100%
• Compression Rate 100%
• Structured Pruning 100%
• Approximation Error 66%
• Worst-Case Guarantees 33%
• Memory Resource 33%
• Limited Computational Resources 33%
• Pruning Strategy 33%
• Comparable Accuracy 33%
• Compression Method 33%
• Compressed Model 33%
• Model Compression 33%

Computer Science

• Neural Network 100%
• Compression Rate 100%
• Approximation (Algorithm) 66%
• Model Compression 33%
• Memory Resource 33%
• Computational Resource 33%

Mathematics

• Neural Network 100%
• Approximation Error 100%
• Approximates 50%
• Worst Case 50%