This repository contains the code and documentation for the Blind Source Separation (BSS) course homework assignments. The assignments cover various topics related to BSS and its applications. Below is a summary of each homework assignment:
-
Part A: BSS problem with linear and instantaneous Sources
-
Part B: Steepest Descend, Newton's Method, Alternation Minimization, Gradient Projection
-
Part A: Principal Component Analysis (PCA) - Singular Value Decomposition (SVD)
-
Part B: Beamforming - Multiple Signal Classification (MUSIC)
-
Part A: Common Spatial Patterns (CSP) combined with Linear Discriminant Analysis (LDA) for motion classification using EEG Signal
-
Part B: Canonical Correlation Analysis (CCA) for identifying the Steady-State Visually Evoked Potential (SSVEP) frequency
-
Part A: BSS poblem with non stationary uncorrelated sources
-
Part B: BSS problem with mutually uncorolated sources
- Sparse recovery problem with different methods:
- Subset selection
- Matching Pursuit (MP)
- Orthogonal Matching Pursuit (OMP)
- Basis Pursuit (BP)
- Iteratively Reweighted Least Square (IRLS)
-
Part A: Least Absolute Shrinkage and Selection Operator (LASSO)
-
Part B: 2d and 3d frame designing
-
Part C: K-SVD and MOD for sparse signal recovery (dictionary learning)
-
BSS problem with independent sources, minimize the Kullback-Leibler divergence (D_kl)
-
BSS problem with independent sources, minimize the Kullback-Leibler divergence (D_kl) with deflation and equivariant methods
-
BSS problem with independent sources, maximize Kurt with fixed-point method (FAST ICA) and another method not sensitive to outliers