Schedule

LECTURE 1

Time: 12:00 - 12:30 

Speaker: Lissa Melis (Department of Quantitative Economics)

Title: ‘Enhancing mobility, efficiency, and transparency through heuristic algorithms’.

Abstract: This presentation focuses on the central role of heuristic algorithms in my current and future research. I use heuristic algorithms across three distinct, yet somewhat interconnected, domains of research: modern logistics and mobility, the statistical analysis of the performance of this type of algorithms, and explainable artificial intelligence (XAI), particularly focusing on counterfactual explanations. 
First, I will demonstrate the essential role of heuristic algorithms in improving urban mobility and logistics efficiency. This will be illustrated through two practical examples: the optimization of on-demand bus transport systems and the strategic placement and setup of automated parcel lockers. Simultaneously, we delve into the statistical analysis of heuristic algorithms, emphasizing the development of robust methodologies for determining optimal stopping points and evaluating algorithmic performance. I will touch on why this research facet is crucial in the field of operations research. Transitioning to the domain of explainable AI, I will introduce the concept of counterfactual explanations, how they often disagree and what are the possible consequences of this disagreement. Also in this area, I will explain how the generation of counterfactual explanation can benefit from the use of heuristic algorithms.

 
LECTURE 2

Time: 12:30 - 13:00 

Speaker: Leto Peel (Department of Data Analytics and Digitalisation)

Title: ‘Statistical inference links data and theory in network science’.

Abstract: The number of network science applications across many different fields has been rapidly increasing. Surprisingly, the development of theory and domain-specific applications often occur in isolation, risking an effective disconnect between theoretical and methodological advances and the way network science is employed in practice. 
Here I will address this risk constructively, discussing good practices to guarantee more successful applications and reproducible results. I will motivate designing statistically grounded methodologies to address challenges in network science. This approach allows one to explain observational data in terms of generative models, naturally deal with intrinsic uncertainties, and strengthen the link between theory and applications.