Learn a new way to approach problem solving by stating the problem and letting powerful constraint solving software do the rest. This class teaches you the art of encoding complex discrete optimization problems in the MiniZinc modeling language and then shows you how to effortlessly solve them by leveraging state-of-the-art open-source constraint solving software.
View the MOOC promotional video here: http://tinyurl.com/huyhq2e
- In this module you will examine some of the archetypal forms of decisions that need to be made in discrete optimization problems and how to represent them in MiniZinc. After this module Sudoku problems will never bother you again.
- In this module you will see how discrete optimization problems can often be seen from multiple viewpoints, and modelled completely differently from each viewpoint. Each viewpoint may have strengths and weaknesses, and indeed the different models can be combined to help each other. You will learn more about converting data into complex constraints or objectives to define a problem. The assignment will challenge you far more than earlier problems.
- You will learn how to encapsulate a complex constraint definition in a predicate definition to enable its reuse. This will enable the construction of far more complex models. You will learn methods to discover what is going wrong with your model and how to fix it. With these tools a complex plannning problem will be easy to solve.
- You will learn how to tackle challenging scheduling and packing problems, and the important combinatorial substructures that underly them. You will see how to model some of the complex constraints that arise in these applications. The assignment will tackle a simplification of a real world combined scheduling and packing problem.