Construction of wavelet dictionaries for ECG modeling

Technical details, algorithms, and MATLAB implementation for a method advanced in the paper ``Wavelet Based Dictionaries Dimensionality Reduction of ECG Signals'', are presented. This work aims to be companion that publication, which an adaptive mathematical model given record is proposed. The comprises following building blocks. (i) Construction suitable redundant set, called 'dictionary', decomposing signal as superposition elementary components, 'atoms', selected from dictionary. (ii) Implementation greedy strategy Optimized Orthogonal Matching Pursuit (OOMP) selecting atoms intervening decomposition. gives details algorithms implementing stage (i), not fully elaborated previous publication. proposed dictionaries constructed known wavelet families, but translating prototypes with shorter step than corresponding basis. Stage readily implementable by available function OOMP. • use software power technique illustrated reducing dimensionality records taken MIT-BIH Arrhythmia Database. has been made publicly on dedicated website. We provide explanations, construction scaling functions 17 different families. procedure designed allow straightforward extension inclusion additional options