Demystifying the ICA black box – Source : Image
This post has been on the back of my mind for quite some time, ever since I wrote about Principal Component Analysis. Independent Component Analysis or ICA is an algorithm to extract underlying structure hidden in multi-dimensional datasets. As for PCA, It is fairly ubiquitous (including in Neuroscience). However, as opposed to PCA, getting an intuitive understanding of ICA is much harder and many persons use it as a magic black box (no offense, I have been there too). I won’t provide comments on any code as fastICA is already freely available in Matlab. My hope here is to facilitate the understanding of what ICA is and how it works. Continue reading
