Syllabus
Every chapter starts from a concrete management problem — how much risk to accept, which price to set, where to locate a service, how smooth a plan should be — turns it into a nonlinear model, solves it with Gurobi called from Python and questions it: how much is one more unit of a resource worth? how much does a smoother plan cost?
Every model can be run right away in the browser: each chapter has its notebook that opens in Colab, with nothing to install. Prerequisites: LP Modelling and Convex Optimization.
The full list of chapters, in the order in which they are covered.
- Introduction to the course — this page and the home page
- Theory: nonlinear optimization — convexity, QP, KKT
- Solver: nonlinear models — functional and bilinear constraints, multipliers
- Production with quadratic smoothing — convex QP
- Supply chain with congestion and CO₂ — convex NLP
- Markowitz portfolio — convex QP
- Pricing and revenue management — NLP
- Advertising budget — convex NLP
- Continuous location — convex NLP
- EV charging with a smooth profile — convex QP
- Queues and service capacity — convex NLP
Optimization and machine learning
- Support Vector Machine — convex QP
- Least-squares regression — convex QP
The course
- Organization of the course — sessions and mistakes to avoid