基于峰態(tài)的獨(dú)立分量分析原理的研究
The Study on the Principle of Kurtosis Based ICA Method
-
摘要: 對(duì)基于峰態(tài)絕對(duì)值最大化的ICA原理進(jìn)行了詳細(xì)分析,給出了該原理幾何解釋和適用范圍。通過(guò)對(duì)采用該原理進(jìn)行ICA問(wèn)題求解過(guò)程的分析,闡明了ICA問(wèn)題解的不確定性的產(chǎn)生原因,指出并解釋了求解過(guò)程中表現(xiàn)出的一種概率特性。所得結(jié)論對(duì)于其它ICA方法也具有參考價(jià)值。
-
關(guān)鍵詞:
- 獨(dú)立分量; ICA; 峰態(tài); 幾何解釋
Abstract: The kurtosis based ICA approach is analyzed particularly and the geometrical explanation of this approach is presented in the paper. Furthermore, we elucidate the reasons of the indeterminacy of ICA solutions and explain the probability property by analyzing the process for finding out the ICA solutions. These analytic results and conclusions are also benefit to the study on other ICA methods. -
Hyvariene A, Oja E. Fast and robust fixed-point algorithms for independent component analysis[J].IEEE Trans. on Neural Networks.1999, 10(3):626-[2]Hyvariene A.[J].Karhunen J, Oja E. Independent Component Analysis [M]. New York: John Wiley Sons Inc.2001,:-[3]Hyvarinen A. Survey on independent component analysis[J].Neural Computing Surveys, 1999, 2:94 - 128.[4]Comon P. Independent component analysis: A new concept?[J].Signal Processing, 1994, 34(4): 287 - 314. -
計(jì)量
- 文章訪(fǎng)問(wèn)數(shù): 2016
- HTML全文瀏覽量: 71
- PDF下載量: 670
- 被引次數(shù): 0
下載:
