ML-for-Biosensing

This repository contains MATLAB scripts for unsupervised analysis of biosignals, specifically EEG and EMG, using time- and frequency-domain features with PCA and k-means clustering.

Contents

1. unsupervised_x1_x2.m:

• detect EMG vs EEG using only x1 and x2 signals

2. unsupervised_with_eeg.m:

• same, plus binned EEG bands (delta, theta, alpha, beta).

Features: time-domain (MAV, RMS, ZCR…), spectral bands, spectral entropy, kurtosis. Analysis: PCA, k-means clustering, signal-level EMG/EEG decision. Outputs: .mat, .csv, and .png plots.

Requirements: MATLAB (compatible with R2014a and later)

Note: Results are saved as .mat and .csv files; plots are saved as .png.