We will end up with much more efficient (low-ranked) matrices for computation based on the original dataset.
The following equation depicts the decomposition of an array of m x n, which is large and hard to work with. The right-hand side of the equation helps to solve the decomposition problem which is the basis of the SVD technique.
The following steps provides a concrete example of the SVD decomposition step by step:
- Consider a matrix of 1,000 x 1,000 which provides 1,000,000 data points (M= users, N = Movies).
- Assume there are 1,000 rows (number of observations) and 1,000 columns (number of movies).
- Let's assume we use ...
