C8591 - PRESCRIPTIVE ANALYTICS

Learning outcomes

This course provides the methods and tools required to address real-world decision-making problems in complex organizational settings. It introduces the mathematical foundations of prescriptive analytics, covering topics from mathematical programming to other relevant methodologies. Through a combination of theoretical lectures and practical sessions, students will learn how to model, solve, and analyze optimization problems that support data-driven decision-making processes. By the end of the course, students will be able to: • Formulate real-world decision-making problems using mathematical programming and other prescriptive analytics techniques. • Apply appropriate optimization methods to support effective and efficient decision-making in complex scenarios. • Use suitable software tools to model and solve prescriptive analytics problems. • Critically analyze and interpret optimization results to derive actionable insights at the strategic, tactical, and operational levels.

Course contents

• Introduction to Prescriptive Analytics:
  • The Analytics Stack
  • Descriptive, Predictive, and Prescriptive Analytics
  • Decision-making lifecycle
  • Business value of optimization
  • Prescriptive Analytics in Digital Transformation
• Mathematical and algorithmic foundations (Optimization, Operations Research):
  • Optimization basics
  • Problem formulation (decision variables, objective functions, constraints)
  • Linear and Integer Programming, sensitivity analysis
  • Nonlinear Programming
  • Heuristics, Metaheuristics, and Matheuristics
  • Decision-Making Under Uncertainty: stochastic optimization, Monte Carlo simulation, Scenario analysis
  • Implementation & Tools: modeling languages, Python (PuLP, Pyomo, OR-Tools), optimization solvers (Gurobi, CPLEX, HiGHS), deployment of decision support systems
• Integration with AI / predictive analytics (without developing AI and predictive analytics methodologies, because they are developed in other courses):
  • Predictive + Prescriptive Analytics
  • Machine Learning for Optimization
  • Optimization for Machine Learning
  • Decision-Focused Learning and Learning-Augmented Optimization
• Applications:
  • Supply chain & logistics
  • Pricing & revenue management
  • Smart manufacturing
  • Resource allocation

Readings/Bibliography

Lecture notes and slides by teacher available online

Further reading on lecture topics:

• C. Papadimitriou, K. Steiglitz, Combinatorial Optimization: Algorithms and Complexity, Dover Publications, NY.
• M.S. Bazaraa, J.J. Jarvis, H.D. Sherali, Linear Programming and Network Flows, Wiley.
• R.K.Ahuja, T.L.Magnanti, J.B.Orlin, "Network flows: theory, algorithms and applications", Prentice Hall.

Teaching methods

Lectures propose both theoretical and practical aspects concerning the different topics. Examples and exercises help students to understand the practical use of the proposed methodologies, templates, and tools.

Assessment methods

A final oral examination assesses the achievement of learning objectives:

• knowledge of the fundamentals of Prescriptive Analytics;
• knowledge of the mathematical methods and algorithms underlying Prescriptive Analytics;
• the ability to apply the approaches studied to real-world decision-making problems.

To be admitted to the final examination, the student must submit a project, previously agreed with the teacher. In preparing the project, students may make limited, declared, and non-substantive use of artificial intelligence (AI) for support activities. Substantive use of AI in carrying out the project is not permitted.

The final assessment consists of an oral examination during which students will discuss their submitted project and demonstrate the knowledge and skills acquired throughout the course.

The final score corresponds to the following level of learning achieved: <18 insufficient; 18-23 sufficient; 24-27 good; 28-30 very good; 30L excellent.

Teaching tools

Lecture notes, slides, seminars, exercises, examples.

Office hours

See the website of Marco Antonio Boschetti