Ruprecht-Karls-Universität Heidelberg




Prüfungsstoff für die Veranstaltung
"Mixed-Integer Programming"

Linear programming

  • Standard LP formulations
  • Duality
  • Polyhedra and basic solutions
  • Primal and dual simplex algorithm
  • Postoptimal analysis and reoptimization
MIP Modelling
  • Decisions, logical conditions, products of binary variables
  • Disjunctions, general integers, semi-continuous variables
  • Partial integers, fixed costs, products of binary and real variables
  • Special ordered sets, piecewise linear functions
  • Incremental price breaks, all items discount
  • Min-max objective functions, ratio objective functions
Combinatorial optimization problems
  • Combinatorial problems
  • Combinatorial problems and integer programming
Computation of optimum solutions
  • Branch-and-bound
  • Cutting plane algorithms
  • Branch-and-cut
  • Dynamic programming
Polyhedral combinatorics
  • Combinatorial polytopes and linear programs
  • Polyhedral theory
  • Separation and optimization
Descriptions of combinatorial polytopes
  • Enumeration
  • Projection
  • Computation techniques
Relaxations
  • Combinatorial relaxation
  • Linear programming relaxation
  • Lagrangian relaxation
Implementation of branch-and-cut algorithms
  • Separation
  • Pricing
  • Fixing and setting of variables
  • Branching
  • Addition of variables
  • Parallelization
Presolve
  • Row and column analysis
  • Coefficient reduction
  • Clique constraints
  • Aggregation and disaggregation
  • Probing
  • Model improvement
Cutting planes
  • Lifting of inequalities
  • Chvatal-Gomory cuts
  • Knapsack cuts
  • Disjunctive cuts
  • Flow cover inequalities
  • Maximally violated mod-k cuts
  • Target cuts
  • Small facet separation
  • Separation for large problems
Column generation
  • Cutting stock problem
  • Dantzig-Wolfe decomposition
  • Benders decomposition


Erstellt am Feb 2, 2011
comopt{at}informatik.uni-heidelberg.de

optWay
Links