论文信息 - A Feature Selection Methodology for Steganalysis

A Feature Selection Methodology for Steganalysis

This paper presents a methodology to select features before training a classifier based on Support Vector Machines (SVM). In this study 23 features presented in [1] are analysed. A feature ranking is performed using a fast classifier called K-Nearest-Neighbours combined with a forward selection. The result of the feature selection is afterward tested on SVM to select the optimal number of features. This method is tested with the Outguess steganographic software and 14 features are selected while keeping the same classification performances. Results confirm that the selected features are efficient for a wide variety of embedding rates. The same methodology is also applied for Steghide and F5 to see if feature selection is possible on these schemes.

[1] Robert Tibshirani,et al. An Introduction to the Bootstrap , 1994 .

[2] Niels Provos,et al. Defending Against Statistical Steganalysis , 2001, USENIX Security Symposium.

[3] Michel Verleysen,et al. Mutual information for the selection of relevant variables in spectrometric nonlinear modelling , 2006, ArXiv.

[4] Nello Cristianini,et al. An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[5] Sayan Mukherjee,et al. Feature Selection for SVMs , 2000, NIPS.

[6] Jessica J. Fridrich,et al. Feature-Based Steganalysis for JPEG Images and Its Implications for Future Design of Steganographic Schemes , 2004, Information Hiding.

[7] Tomás Pevný,et al. Towards Multi-class Blind Steganalyzer for JPEG Images , 2005, IWDW.

[8] Jean-Marc Chassery,et al. Improving LSB steganalysis using marginal and joint probabilistic distributions , 2004, MM&Sec '04.

[9] Sorina Dumitrescu,et al. Detection of LSB steganography via sample pair analysis , 2002, IEEE Trans. Signal Process..

[10] Siwei Lyu,et al. Detecting Hidden Messages Using Higher-Order Statistics and Support Vector Machines , 2002, Information Hiding.