A Hybrid Approach for Optimizing Parallel Clustering Throughput using the GPU

Keyphrases

• Hybrid Method 100%
• Parallel Clustering 100%
• Density-based Spatial Clustering of Applications with Noise (DBSCAN) 100%
• Response Time 50%
• Data Transfer 50%
• Negative Performance 25%
• Between-host 25%
• Galaxies 25%
• Data Distribution 25%
• Dense Datasets 25%
• Search Performance 25%
• Sloan Digital Sky Survey 25%
• Performance Model 25%
• Sparse Data 25%
• Impact Performance 25%
• Main Idea 25%
• GPU Kernel 25%
• Memory Bandwidth 25%
• Reference Implementation 25%
• PCIe 25%
• GPU Memory 25%
• Grid-based 25%
• Ionospheric Total Electron Content 25%
• Buffer Overflow 25%
• Indexing Scheme 25%
• Search Distance 25%
• Concurrent Operation 25%
• Fast Index 25%
• Neighborhood Search 25%
• Time Performance 25%

Computer Science

• Response Time 100%
• Hybrid Approach 100%
• Performance Model 50%
• Memory Bandwidth 50%
• Reference Implementation 50%
• Buffer Overflow 50%
• Performance Impact 50%
• Data Distribution 50%
• Search Neighborhood 50%
• Application Scenario 50%
• Time Performance 50%