... | ... | @@ -136,6 +136,20 @@ We have already discussed a component analysis method, PCA. At this point, you m |
|
|
|
|
|
<img src="uploads/876dc45febab4a5593bcce12383aa455/ica_1.png" width="600">
|
|
|
|
|
|
In PCA, the way we extract the unit vectors of the new coordinate system relies on the variance. If we apply it to a problem composed on two independent phenomena, it will lead to a merged transformation, which is definitely wrong. In such cases, we need aim to filter out the these "independent" behaviors within the data.
|
|
|
|
|
|
|
|
|
Note: In ICA, we assume that we do not have Gaussian distributions in the variables.
|
|
|
|
|
|
Interesting material on ICA:
|
|
|
|
|
|
* [Separating reflections from images by use of independent component analysis](https://www.osapublishing.org/josaa/fulltext.cfm?uri=josaa-16-9-2136&id=1263)
|
|
|
|
|
|
* [Obstruction-Free Photography](https://sites.google.com/site/obstructionfreephotography/)
|
|
|
|
|
|
*[Artefact Correction with ICA](https://towardsdatascience.com/artefact-correction-with-ica-53afb63ad300)
|
|
|
|
|
|
* [Independent component analysis of ocular artifacts](https://hss-opus.ub.rub.de/opus4/frontdoor/deliver/index/docId/2223/file/diss.pdf)
|
|
|
|
|
|
...
|
|
|
|
... | ... | |