独立元的实时算法分析

Real-Time Algorithm of Independent Component

分析了独立元分析(ICA)的多种方法,并应用在声音信号处理中,对结果进行比较和讨论。最后提出实时算法的概念,并对负熵算法进行改进。实验证明,实时负熵算法不仅复杂度更低,而且能自适应学习,具有应用价值。

The paper analyzes the multiple ways of Independent Component Analysis, applies them in the process of voicewav, compares and discusses the result. Finally, it puts forward the concept of Real-Time Algorithm and improves the Engentropy Algorithm. It turns out that Real-Time Engentropy Algorithm can not only lower the complexity, but also make study self-adaptively, so it’s worth in actual application.