This course is designed to provide software professionals with a powerful new tool for handling complex optimization problems.

Target Audience

This course is specifically tailored for software engineers who are proficient in at least one programming language but have little to no background in discrete optimization or operations research. It leverages existing logical thinking to transition students from a procedural "if-this-then-that" mindset to a declarative one focused on modeling desired outcomes. Knowledge of at least one programming language is a prerequisite to successfully engage with the course material and implementation examples.

Course Objectives and Learning Outcomes

By the end of this course, students will be able to:

- Identify Optimization Opportunities: Recognize "hard" problems in their work or daily life that are better suited for optimization than traditional greedy algorithms or heuristics.

- Master Fundamental Modeling: Clearly distinguish and define decision variables, constraints, and objective functions within a mathematical framework.

- Implement Advanced Techniques: Use binary logic to model "exactly one" scenarios, handle logical implications with "Big-M" formulations, and linearize non-linear functions like absolute values and minimums.

- Build Industry-Standard Models: Create MILP models for a wide variety of classic problems.

- Utilize Professional Tools: Implement and solve these models using commercial or open-source solvers, treating them as efficient "black-box super-compilers".

- Refine Model Performance: Understand and apply strengthening techniques, such as clique inequalities and stronger relaxations, to improve solver performance