What is Matrix Factorization?

Missing or unknown data is estimated using a low$-$rank matrix of the known data

Maintaining uniformity in information and leaving data sparse

It is the dot product of the inverse of the matrix and an identity matrix

None of the above