Violeta Teodora Trifunov, M.Sc.
Curriculum Vitae
2017 – 2023 PhD student
Computer Vision Group, Friedrich Schiller University Jena
Climate Informatics Group, German Aerospace Center (DLR), Institute for Data Science, Jena
Research topic: “Deep graphical models and domain knowledge integration”
2015 – 2017 M.Sc. in Mathematics
Rheinische-Friedrich-Wilhelms University Bonn
Master Thesis: “Endomorphism Algebras of Generators-Cogenerators Associated with the Cartan Matrix”
2012 – 2015 B.Sc. in Mathematics
University of Novi Sad
Research Interests
  • Deep Learning
  • Causal Graphical Models
  • Knowledge Integration
  • Causality
  • Anomaly Detection
  • Climate Informatics
Projects
Deep graphical models and domain knowledge integration
Climate data has been vastly accumulated over the past several years, making climate science one of the most data-rich domains. Despite the abundance of data to process, data science has not had a lot of impact on climate research so far, partly due to the fact that ample expert knowledge is rarely exploited. The main goal of this project is bridging the gap between deep learning and causal graphical models while using domain knowledge which could prove to be of significant importance for facilitating an understanding of the Earth system. We aim to develop a sequential version of the Causal Effect Variational Auto-Encoder (CEVAE) and apply it to time series of ecological or climate variables having suitable underlying causal graph structure. When this is accomplished, we intend to apply our method to time series anomaly detection, as well as to variables having more general causal structures.
Publications
2022
Violeta Teodora Trifunov, Maha Shadaydeh, Joachim Denzler:
Sequential Causal Effect Variational Autoencoder: Time Series Causal Link Estimation under Hidden Confounding.
arXiv preprint arXiv:2209.11497. 2022.
[bibtex] [web] [doi] [abstract]
Violeta Teodora Trifunov, Maha Shadaydeh, Joachim Denzler:
Time Series Causal Link Estimation under Hidden Confounding using Knockoff Interventions.
NeurIPS Workshop on A Causal View on Dynamical Systems (NeurIPS-WS). 2022.
[bibtex] [pdf] [web] [abstract]
2021
Violeta Teodora Trifunov, Maha Shadaydeh, Björn Barz, Joachim Denzler:
Anomaly Attribution of Multivariate Time Series using Counterfactual Reasoning.
IEEE International Conference on Machine Learning and Applications (ICMLA). Pages 166-172. 2021.
[bibtex] [pdf] [web] [doi] [abstract]
Violeta Teodora Trifunov, Maha Shadaydeh, Jakob Runge, Markus Reichstein, Joachim Denzler:
A Data-Driven Approach to Partitioning Net Ecosystem Exchange Using a Deep State Space Model.
IEEE Access. 9 : pp. 107873-107883. 2021.
[bibtex] [abstract]
2019
Violeta Teodora Trifunov, Maha Shadaydeh, Jakob Runge, Veronika Eyring, Markus Reichstein, Joachim Denzler:
Causal Link Estimation under Hidden Confounding in Ecological Time Series.
International Workshop on Climate Informatics (CI). 2019.
[bibtex] [pdf] [abstract]
Violeta Teodora Trifunov, Maha Shadaydeh, Jakob Runge, Veronika Eyring, Markus Reichstein, Joachim Denzler:
Nonlinear Causal Link Estimation under Hidden Confounding with an Application to Time-Series Anomaly Detection.
DAGM German Conference on Pattern Recognition (DAGM-GCPR). Pages 261-273. 2019.
[bibtex] [pdf] [abstract]
2018
Violeta Teodora Trifunov, Maha Shadaydeh, Jakob Runge, Veronika Eyring, Markus Reichstein, Joachim Denzler:
Domain knowledge integration for causality analysis of carbon-cycle variables.
American Geophysical Union Fall Meeting (AGU): Abstract + Poster Presentation. 2018.
[bibtex] [web] [abstract]