This project involves processing and analyzing EEG signals using various methods such as GEVD, DSS, PCA, and ICA. The focus is on signal extraction, noise removal, and comparison of different techniques.

• Data: mat1.Q (8-channel EEG signal, 100 seconds, 100 Hz sampling rate)
  • a) Estimate and plot the triangular signal (s1) with a known period of 4 seconds. Calculate and report the estimation error.
  • b) Estimate s1 when the exact period is unknown but within the range of 3 to 7 seconds. Calculate and report the estimation error.
  • c) Estimate s2 using known on/off times (1T). Calculate and report the estimation error.
  • d) Estimate s2 with partial on times provided (2T). Calculate and report the estimation error.
  • e) Estimate s3 with a known frequency band of 10-15 Hz. Calculate and report the estimation error.
  • f) Estimate s3 with a frequency band range of 5-25 Hz. Calculate and report the estimation error.

• Data: Noisy and clean signals from Exercise 2 of Series 2
  • a) Determine spike times from the initial clean signal and store them in a vector.
  • b) Use GEVD and DSS to extract spike sources. Reconstruct and denoise the signals (den_X).
  • c) Plot denoised signals for channels 13 and 24 alongside the original and noisy signals.
  • d) Compute and report RRMSE (Relative RMSE) for each method, noise type, and SNR.
  • e) Compare the results with PCA and ICA from Series 2.

• MATLAB