antonvalk / matrix-vector-equation-neural-network-solution-approximation Goto Github PK
View Code? Open in Web Editor NEWA neural network implementation that predicts what the solution to a sparse matrix equation will be. The prediction is then fed into a solver (using methods such as Gauss-Seidel, Jacobi, etc.) as the initial guess solution vector. The initial guess is close to the solution and therefore the amount of iterations required for convergence is reduced.
License: Apache License 2.0