Roman Abramov

Position

Research assistant, PhD student

E-mail

roman.abramov@tum.de

Address

Room A2.306 / A2.307 (3rd Floor)
Richard-Wagner-Straße 1
80333 München

Office hours

By appointment

 

About Me

I'm currently working towards a Ph.D. degree with the guidance of Prof. Dr. Gjergji Kasneci. My Bachelor's thesis focused on implementing various NLP techniques to verify the authorship of papers in a dead language. During my Master's studies, I delved into Machine Learning in the Quantum Field, where I used Graph Neural Networks to predict the Quantum Random Walk's advantage over classical ones on random graphs. In addition to my academic pursuits, I've worked as a Data Scientist at a medical company. I mostly worked on tabular data classification, such as CVD, diabetes, and some chronic diseases. But a few projects included time-series analysis developing a model to predict COVID-19 dynamics. After that, I worked as a machine learning engineer for a year.

Research Focus

My research interests lie in NLP, memory in Machine Learning, and Graph Neural Networks. I believe that explicit external memory storage for ML models not only boosts their performance but also provides a convenient means for interpretation and fairness in AI.