13. Feature Selection and Evaluation

  1. Hans Georg Schaathun

Published Online: 22 AUG 2012

DOI: 10.1002/9781118437957.ch13

Machine Learning in Image Steganalysis

Machine Learning in Image Steganalysis

How to Cite

Schaathun, H. G. (2012) Feature Selection and Evaluation, in Machine Learning in Image Steganalysis, John Wiley & Sons, Ltd, Chichester, UK. doi: 10.1002/9781118437957.ch13

Author Information

  1. Ålesund University College, Norway

Publication History

  1. Published Online: 22 AUG 2012
  2. Published Print: 21 AUG 2012

ISBN Information

Print ISBN: 9780470663059

Online ISBN: 9781118437957



  • feature selection/evaluation, applications in steganalysis;
  • right features for classification, in pattern matching/machine learning;
  • searching the perfect vector, an NP-complete problem;
  • optimal classification accuracy, model and right level of complexity;
  • model complexity and factors, feature space dimensionality;
  • classifiers in making decision, and curse of dimensionality;
  • selection using information theory;
  • feature selection using information theory, needing multivariate mutual;
  • features and binning, with discrete entropy/mutual information;
  • boosting for feature selection using BFS


This chapter contains sections titled:

  • Overfitting and Underfitting

  • Scalar Feature Selection

  • Feature Subset Selection

  • Selection Using Information Theory

  • Boosting Feature Selection

  • Applications in Steganalysis