Applications of Blind Source Separation Based on Independent Factor Analysis
Liu Ai-lun · Huadong Li-Gong Daxue xuebao · 2008
Most of the existed algorithms on BSS(Blind Signal Separation) are based on ICA(Independent Component Analysis).However,ICA has many limits in actual use.In order to solve this(problem,) a new BSS algorithm based on IFA(Independent Factor Analysis) was proposed in this paper.IFA generalizes and unifies ordinary factor,principal component analysis,and ICA.It is proved by the simulation result that IFA can handle the case that the number of mixtures differs from the number of sources and the data include strong noise.The lower is SNR of data,the better is the predominance of IFA.