Neural Computation
Blind source separation-semiparametric statistical approach
IEEE Transactions on Signal Processing
Efficient source adaptivity in independent component analysis
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
We describe an efficient method for accurate estimation of the score function of a random variable, which can be regarded as an extension of the FFT-based fast density estimation method of Silverman (1982), and which scales no more than linearly with the sample size. We demonstrate the utility of our approach in a real-life ICA problem involving the separation of eight sound signals, where better results are observed than using state-of-the-art ICA methods.