A New Learning Algorithm for Blind Signal Separation
Citations
17,362 citations
14,635 citations
Additional excerpts
...Many UL methods are designed to maximize information-theoretic objectives (e.g., Linsker, 1988; Barlow et al., 1989; MacKay and Miller, 1990; Plumbley, 1991; Schmidhuber, 1992b,c; Schraudolph and Sejnowski, 1993; Redlich, 1993; Zemel, 1993; Zemel and Hinton, 1994; Field, 1994; Hinton et al., 1995; Dayan and Zemel, 1995; Amari et al., 1996; Deco and Parra, 1997), and to uncover and disentangle hidden underlying...
[...]
10,141 citations
8,231 citations
Cites methods from "A New Learning Algorithm for Blind ..."
...The above version of FastICA could be compared with the stochastic gradient method for maximizing likelihood ( Amari et al., 1996; Bell and Sejnowski, 1995; Cardoso and Laheld, 1996; Cichocki and Unbehauen, 1996):...
[...]
...Finally, we give a version of FastICA that shows explicitly the connection to the well-known infomax or maximum likelihood algorithm introduced in ( Amari et al., 1996; Bell and Sejnowski, 1995; Cardoso and Laheld, 1996; Cichocki and Unbehauen, 1996)....
[...]
8,091 citations
Additional excerpts
...Further reading on blind separation, including non-ICA algorithms, can be found in (Jutten and Herault, 1991; Comon et al., 1991; Hendin et al., 1994; Amari et al., 1996; Hojen-Sorensen et al., 2002)....
[...]
References
9,157 citations
"A New Learning Algorithm for Blind ..." refers background or methods in this paper
...Although the on-line learning algorithms (16) and (19) look similar to those in [3, 7] and [5] respectively, the selection of the activation function in this paper is rational, not ad hoc....
[...]
...Several neural network algorithms [3, 5, 7] have been proposed for solving this problem....
[...]
...It is a non-monotonic activation function different from those used in [3, 5, 7]....
[...]
8,522 citations
"A New Learning Algorithm for Blind ..." refers background or methods in this paper
...The minimization of the Kullback-Leibler divergence leads to an ICA algorithm for estimating W in [6] where the Edgeworth expansion is used to evaluate the negentropy....
[...]
...In practice, other activation functions such as those proposed in [2]-[6] may also be used in (19)....
[...]
...The algorithm in [6] is based on the Edgeworth expansion[8] for evaluating the marginal negentropy....
[...]
...The mathematical framework for the ICA is formulated in [6]....
[...]
...Different from the work in [6], we use the Gram-Charlier expansion instead of the Edgeworth expansion to calculate the marginal entropy in evaluating the MI....
[...]
2,583 citations
"A New Learning Algorithm for Blind ..." refers background or methods in this paper
...Although the on-line learning algorithms (16) and (19) look similar to those in [3, 7] and [5] respectively, the selection of the activation function in this paper is rational, not ad hoc....
[...]
...Several neural network algorithms [3, 5, 7] have been proposed for solving this problem....
[...]
...How should the activation function be determined to minimize the MI? Is it necessary to use monotonic activation functions for blind signal separation? In this paper, we shall answer these questions and give an on-line learning algorithm which uses a non-monotonic activation function selected by the independent component analysis (ICA) [7]....
[...]
...It is a non-monotonic activation function different from those used in [3, 5, 7]....
[...]
1,417 citations
"A New Learning Algorithm for Blind ..." refers methods in this paper
...which has the same "equivariant" property as the algorithms developed in [4, 5]....
[...]