Non-linear independent component analysis with diffusion maps
نویسندگان
چکیده
منابع مشابه
Non Linear Independent Component Analysis with Diffusion Maps
We introduce intrinsic, nonlinearly invariant, parameterizations of empirical data, generated by a nonlinear transformation of independent variables. This is achieved through anisotropic diffusion kernels on observable data manifolds that approximate a Laplacian on the inaccessible independent variable domain. The key idea is a symmetrized second order approximation of the unknown distances in ...
متن کاملLocally linear independent component analysis
Linear Independent Component Analysis (ICA) has become an important technique in unsupervised neural learning. Even though linear ICA yields meaningful results in many cases, it can provide a crude approximation only for general nonlinear data distributions. In this paper we study techniques where local ICA models are applied to data rst grouped or clustered using some suitable algorithm. The g...
متن کاملNon-linear Independent Component Analysis for Speech Recognition
This paper addresses the problem of representing the speech signal using a set of features that are approximately statistically independent. This statistical independence simplifies building probabilistic models based on these features that can be used in applications like speech recognition. Since there is no evidence that the speech signal is a linear combination of separate factors or source...
متن کاملRank based Least-squares Independent Component Analysis
In this paper, we propose a nonparametric rank-based alternative to the least-squares independent component analysis algorithm developed. The basic idea is to estimate the squared-loss mutual information, which used as the objective function of the algorithm, based on its copula density version. Therefore, no marginal densities have to be estimated. We provide empirical evaluation of th...
متن کاملNonlinear Independent Component Analysis by Self-Organizing Maps
Linear Independent Component Analysis considers the problem of nd-ing a linear transformation that makes the components of the output vector statistically independent. This can be applied to blind source separation, where the input data consist of unknown linear mixtures of unknown independent source signals. The original source signals can be recovered from their mixtures using the assumption ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Applied and Computational Harmonic Analysis
سال: 2008
ISSN: 1063-5203
DOI: 10.1016/j.acha.2007.11.001