Selin Aviyente

Selin Aviyente is a Professor in the Department of Electrical and Computer Engineering at Michigan State University's College of Engineering. Her research focuses on statistical signal processing, machine learning, and computational neuroscience, with emphasis on non-stationary signal analysis, tensor decompositions, and community detection in multi-layer networks. She has developed methods for robust tensor decomposition, graph signal processing, and applications to functional connectivity in the brain.

• Ph.D., Electrical Engineering, University of Michigan (2002)
• M.S., Electrical Engineering, University of Michigan (1999)
• B.S., Electrical Engineering, Bogazici University (1997)

Her research integrates tensor-based methods for analyzing high-dimensional biomedical data, with a focus on dynamic functional connectivity and phase-amplitude coupling in neuronal oscillations. She has pioneered frameworks for multiview graph learning, signed graph inference, and spatiotemporal anomaly detection.

The most recent publications highlight trends in tensor decompositions (e.g., robust and higher-order PCA), community detection in multiplex/hypergraphs, graph signal processing for brain networks, and multiview learning applied to neuroimaging and single-cell genomics. Key subtopics include hub node identification, discriminative subgraph detection, and spatiotemporal modeling.

Scientific Awards:

• Withrow Teaching Excellence Award
• NSF Career Award (2008)

Dr. Aviyente advises Ph.D. students and undergraduates in signal processing and computational neuroscience. Her group has secured multiple NSF grants, including CIF: Small: Multiview Graph Learning (2022-present) and CIF: Small: Community Detection in Multilayer Networks (2020-2023). Current projects include NSF-SNSF collaborations for biomedical graph representations.

She leads the Signal Processing Lab (SPLab-aviyente) at MSU, which develops open-source tools for graph learning, tensor analysis, and community detection. The lab's GitHub repositories include PyGraphLearning, dynSGL, and GRASCale, focusing on Python/MATLAB implementations for biomedical and traffic data applications.