# Independent Component Analysis

> Source: https://sukruyusufkaya.com/en/glossary/independent-component-analysis
> Updated: 2026-05-13T21:00:56.205Z
> Type: glossary
> Category: makine-ogrenmesi
**TLDR:** A dimensionality reduction and separation method that aims to decompose mixed signals into statistically independent components.

<p>Independent Component Analysis is especially useful for separating mixed signals and identifying hidden independent sources. Examples include EEG signals, audio mixtures, and some financial time-series settings. Unlike PCA, it does not focus only on variance, but on statistical independence. This makes it a powerful alternative for signal separation and latent structure discovery.</p>