A Minimum Angle Method for Dual Feasibility

TL;DR

This paper introduces a new minimum angle method for efficiently finding dual feasible bases in linear programming, demonstrating superior speed over classical pivot rules through experimental comparison.

Contribution

The paper proposes a novel minimum angle rule approach that is simpler and faster than existing methods for dual feasibility in linear programming.

Findings

The new method outperforms classical pivot rules in speed.

Experimental results confirm the efficiency of the proposed approach.

The approach simplifies the process of finding dual feasible bases.

Abstract

In this paper we presented a new driving variable approach in minimum angle rule which is simple and comparatively fast for providing a dual feasible basis. We also present experimental results that compare the speed of the minimum angle rule to the classical methods. The experimental results showed that this algorithm outperforms all the previous pivot rules.