Group of Inneke Van Nieuwenhuyse

Prof. Van Nieuwenhuyse focuses her research on solving noisy multiobjective optimization problems with scarce data. These occur commonly, e.g. in engineering and in business . In most cases, the objectives need to be observed through (potentially time consuming) physical or computer experiments. Her team develops efficient and effective methods to solve such problems, by combining expertise from the computer science field (Gaussian Process Regression), and stochastic operations research (modelling and optimization of systems with heterogeneous noise). She is now working as a professor at DSI.

Alejandro Morales Hernández's doctoral research is entitled “Optimization of machine learning models: applications in forecasting”. Within this topic, the focus is developing a data-efficient multi-objective Hyperparameter Optimization (HPO) method for machine learning algorithms, which enables the analyst to automatically find the Pareto-optimal solution configurations, considering the uncertainty in the corresponding performance metrics.