• LO1: To provide basic concepts of multivariate signals
• LO2: To impart knowledge on statistical analysis of multivariate time series data
• LO3: To introduce time and spectral domain approaches for analysing multivariate biomedical data

• CO1: Ability to understand the basics of multivariate signal processing
• CO2: Ability to apply statistical analysis for multivariate time series data
• CO3: Ability to analyse multi-domain features of Biomedical signals
• CO4: Ability to evaluate performance of multivariate signal processing algorithms

Concept of random variables – Stochastic processes – Relations among random variables – correlation, multiple correlation, and partial correlation – Univariate and multivariate Gaussian distributions – Univariate Time Series – Time domain approach – Frequency domain approach.

Time series models – AR Models, ARMA Models – Multivariate Time Series – Time domain approach and spectral domain approach – Assessing relations among time series in the spectral domain – Data based estimation versus model based estimation – Principal Component Analysis (PCA) – Signal decorrelation – Independent Component Analysis (ICA).

Data compression of EEG and ECG signals – EMG Source signal separation techniques – EEG signal separation and Pattern Classification – Correlation of Biomedical signals – Evaluating causal relations in biomedical systems – Case studies – ICA based analysis on neurological disorders using EEG – Deep learning-based arrhythmia classification using EEG.

1. William W. S. Wei, Multivariate Time Series Analysis and Applications, Wiley, 2019
2. Katarzyn Blinowska, Jaroslaw Zygierewicz, Practical Biomedical Signal Analysis Using MATLAB -Multiple channels (multivariate) signal, CRC press, 2011.
3. Johnson, Applied Multivariate Statistical Analysis, PHI publisher, 2012.
4. Jocelyn Chanussot, Jocelyn Chanussot, Kacem Chehdi, Multivariate Image processing, Wile Publication, 2009.