Technical report , August 2011 Ensemble classification in steganalysis – Cross-validation and AdaBoost Ensemble classification in steganalysis – Cross-validation and AdaBoost

Two alternative designs to the ensemble classifier proposed in [13] are studied in this report. First, the out-of-bag error estimation is replaced with crossvalidation. Second, we incorporate AdaBoost and modify the weights of the individual training samples as the training progresses. The final decision is formed as a weighted combination of individual predictions rather than through majority voting. We experimentally compare both alternatives with the original design and conclude that they bring no performance gain.

[1]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1995, EuroCOLT.

[2]  Tin Kam Ho,et al.  The Random Subspace Method for Constructing Decision Forests , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  David G. Stork,et al.  Pattern classification, 2nd Edition , 2000 .

[4]  Francis K. H. Quek,et al.  Attribute bagging: improving accuracy of classifier ensembles by using random feature subsets , 2003, Pattern Recognit..

[5]  Phil Sallee,et al.  Model-Based Steganography , 2003, IWDW.

[6]  Robert E. Schapire,et al.  The Boosting Approach to Machine Learning An Overview , 2003 .

[7]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[8]  D. Ruppert.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction , 2004 .

[9]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[10]  Tomás Pevný,et al.  Statistically undetectable jpeg steganography: dead ends challenges, and opportunities , 2007, MM&Sec.

[11]  James J. Chen,et al.  Classification by ensembles from random partitions of high-dimensional data , 2007, Comput. Stat. Data Anal..

[12]  Anindya Sarkar,et al.  Further study on YASS: steganography based on randomized embedding to resist blind steganalysis , 2008, Electronic Imaging.

[13]  Zhiwen Yu,et al.  Image classification based on the bagging-adaboost ensemble , 2008, 2008 IEEE International Conference on Multimedia and Expo.

[14]  Ivan Laptev,et al.  Improving object detection with boosted histograms , 2009, Image Vis. Comput..

[15]  Gürsel Serpen,et al.  Classification in High-Dimensional Feature Spaces: Random Subsample Ensemble , 2009, 2009 International Conference on Machine Learning and Applications.

[16]  Jessica J. Fridrich,et al.  Calibration revisited , 2009, MM&Sec '09.

[17]  Tomás Pevný,et al.  Modern steganalysis can detect YASS , 2010, Electronic Imaging.

[18]  Matthias O. Franz,et al.  Simple algorithmic modifications for improving blind steganalysis performance , 2010, MM&Sec '10.

[19]  Jan Kodovský.  On dangers of cross-validation in steganalysis On dangers of cross-validation in steganalysis , 2011 .

[20]  Jessica J. Fridrich,et al.  Steganalysis in high dimensions: fusing classifiers built on random subspaces , 2011, Electronic Imaging.

[21]  Jessica J. Fridrich,et al.  Ensemble Classifiers for Steganalysis of Digital Media , 2012, IEEE Transactions on Information Forensics and Security.