M. S. Bennet Praba

Adaptive Dynamic Feature Selection using Evolutionary Meta-Learning for Streaming Data

This paper presents an Evolutionary Meta-Learning Framework (EMLF) for dynamic feature selection in high-velocity, non-stationary data streams. Unlike static methods, EMLF continuously updates the selected feature subset as feature relevance changes over time. A multi-objective fitness function, normalized using predic...