Pitfalls and best practices in parametric identification

Mathematical models commonly used in science and engineering are usually provided in terms of Ordinary/Partial Differential Equations (ODEs/PDEs) or Difference Equations. The parameters of these equations are not always exactly known, and they have to be retrieved from experiments. This seminar will convey the most used methods for data-driven parameter estimation, highlighting, through the use of real-world examples, advantages and pitfalls, and proving suggestions for practical implementation and debugging.