Skip to content

Principal Component Analysis

Sambit Paul edited this page Dec 2, 2023 · 6 revisions

We apply the Principal Component Analysis on multi-dimensional data to reduce the dimensions while retaining maximal information by determining which features have maximum variance.

The examples provided here use triaxial raw data collected from an accelerometer and smoothed:

signal

CODE
PCA p1 = new PCA(raw_signal, 1); // raw_signal of type double, output will be of 1 dimension
p1.fit();
double[][] output = p1.transform();

PCA OUTPUT

dft

Clone this wiki locally