K-SVD

K-SVD stands for K-Singular Value Decomposition. It is an algorithm that helps to find a way to represent data using fewer numbers (called vectors). It takes the data that you give it and breaks it down into simpler pieces which can be used to represent the same information with less memory. K-SVD finds the best way of breaking the data down into simpler pieces, so that it is still able to capture all of the information. It does this by finding patterns inside the data and using those patterns to create simpler and smaller vectors.