Data-Adaptive Active Sampling for Efficient Graph-Cognizant Classification

Keyphrases

• Sampling Methods 100%
• Data-adaptive 100%
• Sampling Strategy 100%
• Markov Random Field 100%
• Adaptive Active Sampling 100%
• Computationally Efficient 50%
• Large Graphs 50%
• Uncertainty Sampling 50%
• Training Data 50%
• Numerical Test 50%
• Sample Variance 50%
• Binary Classification 50%
• Categorical Model 50%
• Similarity Measure 50%
• Neighbor Node 50%
• Combinatorial Complexity 50%
• Optimality Criteria 50%
• Active Sampling 50%
• Continuous Values 50%
• Nodal Features 50%
• Variance Minimization 50%
• Model Retraining 50%
• Label Prediction 50%
• Exploration Capability 50%

Mathematics

• Markov Random Fields 100%
• Training Data 50%
• Binary Classification 50%
• Synthetic Data 50%
• Gaussian Distribution 50%
• Sampling Variance 50%
• Similarity Measure 50%
• Optimality Criterion 50%
• Real Data 50%

Computer Science

• Approximation (Algorithm) 100%
• Training Data 100%
• Binary Classification 100%
• Neighboring Node 100%