# Second-order Approximation of Minimum Discrimination Information in Independent Component Analysis

arXiv: https://arxiv.org/abs/2111.15060 · HTML: https://mlsanity.com/paper/2111.15060/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 3 of its 33 references that are in the map.
- First posted on arXiv 30 Nov 2021.

## Builds on (most cited first, 3 of 3)

- [Faster independent component analysis by preconditioning with Hessian approximations](https://mlsanity.com/paper/1706.08171/) (2017)
- [Overcomplete Independent Component Analysis via SDP](https://mlsanity.com/paper/1901.08334/) (2019)
- [Independent Component Analysis via Distance Covariance](https://mlsanity.com/paper/1306.4911/) (2013)

## Built on by (most cited first, 1 of 1)

- [Boosting Independent Component Analysis](https://mlsanity.com/paper/2112.06920/) (2021)

## Abstract (arXiv)

Independent Component Analysis (ICA) is intended to recover the mutually independent sources from their linear mixtures, and F astICA is one of the most successful ICA algorithms. Although it seems reasonable to improve the performance of F astICA by introducing more nonlinear functions to the negentropy estimation, the original fixed-point method (approximate Newton method) in F astICA degenerates under this circumstance. To alleviate this problem, we propose a novel method based on the second-order approximation of minimum discrimination information (MDI). The joint maximization in our method is consisted of minimizing single weighted least squares and seeking unmixing matrix by the fixed-point method. Experimental results validate its efficiency compared with other popular ICA algorithms.