One of the main problems connected with the formal concept analysis and lattice construction is the high complexity of algorithms which plays a significant role when computing all concepts from a huge incidence matrix. In some cases, we only need to compute some of them to test for common attributes. In our research we try to modify an incidence matrix using matrix decompositions, creating a new matrix with fewer dimensions as an input for some known algorithms for lattice construction. In this paper, we want to describe methods of matrix decompositions and their influence on the concept lattice.