Solution to Linear Programming Patterns using Machine Learning

In this research work, linear programming patterns with bounded variables are solved by Supervised Neural Network with back propagation (SNN). A decomposition of large scale optimization tasks into lower order tasks, in optimization problem, is proposed. Samples of linear programming problems (LPPs) are converted into linear programming patterns. The SNN is able to give correct solution to pattern classification. The novel approach of formulation of the linear programming pattern is important in this paradigm. The detailed experimental results are reported.