About
I am a research scientist at Accenture Labs, Dublin. I recently obtained my PhD at Swiss Federal Institute of Technology, Lausanne (EPFL). My doctoral thesis was on representation learning for multi-relational data where my thesis director was Pierre Vandergheynst. I like to research machine learning for graph-structured data and I am broadly interested in knowledge representation and reasoning for explainable AI.
I also have a signal processing background, which I majored in MSc. My master thesis was on the analysis of airborne LiDAR point clouds using the graph signal processing framework. I completed it under the co-supervision of Aydin Alatan and Elif Vural.
Interests
- Graph Representation Learning
- Network Science
- Signal Processing on Graphs
- Knowledge Graph Reasoning
Education
-
PhD in Electrical-Electronics Eng., 2021
École Polytechnique Fédérale de Lausanne
-
MSc in Electrical-Electronics Eng., 2017
Middle East Technical University
-
BSc in Electrical-Electronics Eng., 2013
Middle East Technical University
Featured Publications
Propagation on Multi-relational Graphs for Node Regression
Multi-relational Propagation (MrP) suggests a propagation approach designed for CONTINUOUS labels, and how to complete them on multi-relational and directed graphs.
Node Attribute Completion in Knowledge Graphs with Multi-Relational Propagation
In this study, we address the incompleteness in the numerical node attributes of a knowledge graph. We propose the algorithm MrAP, Multi-Relational Attribute Propagation, which we formulate within a message passing scheme.
Mask Combination of Multi-layer Graphs for Global Structure Inference
The structure of the data very often depends on complex relationships of multiple types. In this study, our main research question is “Can we support the structure inference process with a priori relational information about the data domain?”