基于特征融合的加权SVM音频隐写分析算法

Based on feature fusion weighted SVM audio steganographic analysis algorithm

针对隐写分析中特征维数过高的问题,提出一种特征加权支持向量机音频隐写分析算法。利用特征相关性对原始特征进行优化选择,利用增益比率法计算特征权重,提出了改进特征加权支持向量机。与常用的C-SVM进行的对比实验表明,该方法能够有效提高检测率,降低时间复杂度。

in view of the characteristics of high dimension problems, put forward a feature weighted support vector machine （ SVM ） audio steganographic analysis algorithm.Using correlation characteristics of original features optimized choice, using the gain ratio method to calculate weight characteristics, a feature weighted support vector machine （ SVM ） is presented.With the commonly used C - SVM through the contrast experiments show that this method can effectively increase the detection rate, reduce the time complexity.