Overview
Charles Elkan is a computer scientist (University of California, San Diego), known for foundational work on cost-sensitive learning, probability calibration, and applied machine learning.
Key Contributions / Features
- Probability calibration (Zadrozny-Elkan 2002): isotonic-regression calibration and multiclass calibration by binary decomposition and coupling.
- The Foundations of Cost-Sensitive Learning (Elkan 2001): the theory of decision thresholds under example-dependent misclassification costs.
Related