Less is more: optimizing classification performance through feature selection in a very-high-resolution remote sensing object-based urban application

Earth and Planetary Sciences

• Machine Learning 100%
• Land Cover 100%
• Space Storage 100%
• Image Analysis 100%
• Remote Sensing 100%
• Land Use 100%
• Support Vector Machine 100%

Engineering

• Classification Performance 100%
• Feature Extraction 100%
• High Resolution 100%
• Random Forest 50%
• Metrics 33%
• Recursive 33%
• Selection Algorithm 33%
• Classification Accuracy 16%
• Land Use 16%
• Image Analysis 16%
• Selection Method 16%
• Computational Time 16%
• Reward Model 16%
• Processing Time 16%
• Named Selection 16%
• Learning System 16%
• Nearest Neighbor 16%
• Land Cover 16%
• Support Vector Machine 16%

Computer Science

• Feature Selection 100%
• high resolution remote sensing 100%
• Classification Performance 100%
• Feature Extraction 100%
• Random Decision Forest 42%
• Image Analysis 14%
• Correlation Algorithm 14%
• Analysis Framework 14%
• Classification Models 14%
• Extreme Gradient Boosting 14%
• Classification Accuracy 14%
• Support Vector Machine 14%
• Processing Time 14%
• Computational Time 14%
• Named Selection 14%
• Machine Learning 14%
• Learning System 14%