A simplex algorithm for piecewise-linear programming. III: Computational analysis and applications (Q1184346)

[For part II see the author, ibid. 41, No. 3, 281-315 (1988; Zbl 0656.90062).] This is the last part of a paper devoted to a simplex algorithm for minimizing convex separable piecewise-linear functions subject to linear constraints. The author demonstrates the inherent advantages of the direct piecewise-linear simplex implementation over an indirect approach that relies on transformation to a linear program. These advantages are shown to be implicit in relationships between the linear and piecewise- linear algorithms and largely independent of the details of implementation (in two sets of computational results, the piecewise- linear simplex is observed to run 2-6 times faster than a comparable linear algorithm). A survey of many varied applications of piecewise linear programming is also given.