Data Science at the Service of Neurosciences: Meet Eftychia Solea

Eftychia Solea recently joined ENSAI and CREST as a researcher and an assistant professor of Statistics. Born in Cyprus, her academic career took her to the University of Cambridge and Penn State. Currently working on nonparametric graphical models for functional data, she is looking forward to collaborating with ENSAI fellow researchers on further areas. 

Eftychia Solea teaches two classes in the Master’s in Statistics for Smart Data: “Graphical models & dynamic networks” and “Functional data analysis”, with Professor Valentin Patilea, research director at ENSAI.

Creating new methodologies to analyze data with complicated features

Solea’s research interests lie at the intersection of machine learning and functional data analysis. She is particularly interested in developing statistical techniques for analyzing a specific kind of data: those that are massive in size, complicated in structure, and are very common in today’s applications such as neurosciences, for example, fMRI data.

Analyzing this kind of data with classic statistical methods is difficult because they are characterized by complicated features such as non-linearity.

Solea’s research contributes to the creation of new methodologies to analyze this data.

Nonparametric graphical models for functional data

Studying the development of nonparametric graphical models for functional data with application to brain networks based on fMRI is the main focus of Solea’s research.

She is currently working on a research project on this very subject in collaboration with Professor Dr. Holger Dette from the University of Bochum. This work was motivated by brain network connectivity based on fMRI data.

fMRI data are types of brain signals recorded at each region of the brain at different times, measuring brain activity. The researcher works to construct graphical models to characterize brain network connectivity based on these signals. In this case, the nodes of the graph correspond to regions of the brain and edges indicate associations between the brain regions.

“This is important because it allows neuroscientists to see how the regions of the brain interact with each other in serious disorders such as Alzheimer’s disease or schizophrenia”.

Solea first came across this subject while studying for her doctoral degree at Pennsylvania State University. She wrote a paper about function graphical models using Gaussian assumptions with Professor Bing Li. It was also the subject of Eftychia Solea’s Ph.D. thesis, under the direction of Professor Bing Li at Penn State.

She is also currently working on another related subject with Temple University, USA.

The road from Cyprus to ENSAI

Eftychia Solea obtained her Bachelor’s degree in Applied Mathematics at the National Technical University of Athens. She graduated from the University of Cambridge with a Master’s in Mathematics, after which she enrolled at Penn State for her doctoral degree. She then went back to Cyprus where she worked two years as a post-doctoral fellow at the University of Cyprus before starting a second post-doc at Rühr-Universität in Bochum, Germany.

Solea is a world traveler and so is her Cypriot cat, who made the move to Germany and now calls Rennes home too!

The researcher in statistics has enjoyed her first weeks at ENSAI and its green campus. She has started taking French lessons but has appreciated the international environment at school and her helpful English-speaking colleagues.

Beyond nonparametric graphical models for functional data, Solea is interested in differential privacy, which was the subject of her Master’s thesis. The aim is to construct statistical techniques to analyze personal data without compromising the privacy of individuals.

“When I applied at ENSAI, I saw that my research work matched with some of the professors’. I am looking forward to collaborating with my colleagues“.

Find out more about research at ENSAI.