Optimization models and algorithms for innovation

Optimization models and algorithms for innovation

InstructorMichele Fedrizzi
PeriodSecond semester
Length (h)6

Course Objectives

This course provides some optimization methods that can be fruitfully applied on innovations-related problems.

After a short overview on classical optimization techniques, the course will offer an introduction to a set of non-classical optimization methods called Evolutionary Algorithms (EA). EA were introduced in the past decades and are inspired by biological evolution. From a technical point of view, they are a family of population-based trial and error problem solvers. EA use mechanisms such as reproduction, mutation, recombination (crossover) and selection. The most popular EA are called ‘genetic algorithms’.

EA methods appear to be particularly well suited for problems that are hard to solve in real world applications. One of the major advantages of EA methods, compared to other methods, is that they only need little problem specific knowledge and that they can be applied on a broad range of problems. Typical applications of EA are hard combinatorial problems. Finance-related applications of EA include constrained portfolio selection, trading rules, bankruptcy prediction, credit scoring and data mining.

The course will focus on the main ideas of the described methods, without giving full details of the various algorithms.

Course Content Summary

  • Overview on classical optimization techniques.
  • Evolutionary Algorithms: main concepts and mechanisms, reproduction, mutation, recombination (crossover), and selection.
  • Applications.
  • Software and computer programs.

Teaching methods

Lectures and teamwork


Students will be provided with study material during classes.