I am performing binary (modulo 2) operations on binary matrices (containing only 1s or 0s).
In order to invert such matrix using binary arithmetic I am currently converting it to GF(2). The process works well, however Galois Fields take as much memory as single precision real numbers. This leads to 'out-of-memory errors' for large matrices.
Is there an efficient way to invert a large sparse matrix in GF(2)?
Best Answer