Абстрактный

Comprehensive exploration of the independent component based on the study of music signal collection

Decun You

The data analysis needed in collecting music signal can be divided into two kinds: one is the Principal Component Analysis ; the other one is Independent component analysis. In the collecting process both methods all have their own superiority, but the later one has less errors than the former one. So in the exploration, this paper firstly introduce the principal component analysis, whose key part and also one of preconditions of its practice is the pretreatment. After the pretreatmentÂ’s efficient projection and integration to original data and multidimensional data, bleaching process will effectively analysis the linear variation of other vector data, so that intensity projection can keep in highly accurate. Finally this paper systematically presents the theory of Independent Component Arithmetic, and in this process the paper largely explores maximization, no gauss and kurtosis, making Independent Component Analysis and Principal Component Analysis harmoniously develop in this study. By this way two methods can have a positive effect on the increasing rationality of this study and lay a solid method and data foundation for the music signal collection, making the analysis more reasonable and the analysis steps more targeted.