Data Stream Mining

Tackling different challenges in data streams

Data streams are potentially unbounded data sequences that are made available over time. Mining data streams require algorithms that are:

  • Fast
  • Memory-bounded, and
  • Adaptive, as data are potentially ephemeral due to concept drifts

I am currently interested in different aspects regarding data stream mining:

  • Classification
  • Regression
  • Clustering
  • Feature Selection
  • Data representation and concept drift in text data
  • Recommender systems