[Profile] Fehmina Malik, postdoctoral researcher in artificial intelligence (game theory, explainable artificial intelligence (XAI), emotion detection) (UPR LIA)
What is your research about?
My research focuses on applications of game theory in explainable artificial intelligence (XAI), particularly in emotion detection models. I study how Shapley values, inspired by cooperative game theory, contribute to model interpretability. In addition, I explore the impact of feature correlation on SHAP attributions, with the aim of refining explicability methods in machine learning.
What are your current scientific activities?
I have published papers in leading journals such as IEEE Transactions on Networking (IEEE TON) and IEEE Transactions on Network and Service Management (IEEE TNSM), focusing on network economics. My research has also been presented at leading international conferences such as WIOPT, Performance and COMSNET.
Why did you choose to work in academic research?
Academic research allows me to explore fundamental questions in machine learning and optimisation while contributing to practical advances. The interdisciplinary nature of my work - which covers AI, game theory and operations research - motivates me to develop innovative solutions that improve transparency and decision-making in automated systems.
What advice would you give to students who want to do research?
Curiosity and perseverance are the keys to research. It is essential to develop strong analytical skills and to see challenges as learning opportunities. Collaborating across disciplines and keeping abreast of the latest developments can open up new perspectives and enhance the impact of research.
What object or image from your business best illustrates you?
A Rubik's CubeThis image, partially solved to reveal a network diagram, represents the complex puzzle of AI explainability and my work to unravel it with Shapley values and feature correlation. This image bridges the gap between the abstract concepts of game theory and AI and the concrete goal of model transparency.
The Avignon Computer Laboratory (UPR LIA)
The LIA (Laboratoire Informatique d'Avignon) is a Proprietary Research Unit (UPR 4128) that brings together Avignon Université's lecturers and researchers in computer science, as well as administrative and technical staff, engineers, doctoral students and masters students on research placements. The laboratory is part of the Centre d'Enseignement et de Recherche en Informatique (CERI) at Avignon Université.
With a staff of around 90, including 29 permanent employees and around 40 PhD students working in the fields of automatic speech and language processing, networks and operational research, the laboratory is structured into two teams: CORNET & SLG.
Fehmina Malik is a member of the CORNET team. Her postdoctorate is supported by the LIAvignon partnership chair in artificial intelligence.
The portraits
Mis à jour le 6 March 2025