Data Science Day Jena 2023

Session

Poster session

10 May 2023, 13:50

Seminarraum (Rosensäle)

11. ACQuA: Answering Comparative Questions with Arguments

In the ACQuA project, we develop algorithms to understand and answer comparative information needs like "Is a cat or a dog a better friend?" by retrieving and combining facts, opinions, and arguments from web-scale resources. Ideally, an answer explains why and under what circumstances which comparison alternative should be chosen.

Retrieval-based comparative question answering starts with...

12. Automated Provision of Data Science Tools to End-Users using Docker & Kubernetes

Methods of data science, especially those dealing with machine learning are irreplaceable tools in may fields of science today. Without these methods, many experimental or observational results could not be transferred to useful scientific results. However, in many cases the researchers developing data science methods are not the end-users of them. Especially for todays complex models, such as...

5. Causal Discovery using Model Invariance through Knockoff Interventions

Wasim Ahmad (Computer Vision Group, Friedrich Schiller University Jena)

Cause-effect analysis is crucial to understand the
underlying mechanism of a system. We propose
to exploit model invariance through interventions
on the predictors to infer causality in nonlinear
multivariate systems of time series. We model
nonlinear interactions in time series using
DeepAR and then expose the model to different
environments using Knockoffs-based interven-
tions to...

21. Data and its challenges on the path to end-to-end digitization in public administration - Contributions from three projects of the openDVA working group

The implementation of the right to digital access (OZG) in Germany stops at the office door focusing only on the needs of citizens. It does not cover any internal administrative processes and leaves out various stakeholders. For true end-to-end digitization, we need detailed, interoperable descriptions that can be exploited by all interested parties, including small to medium enterprises,...

14. Epidemiology of genomic surveillance data: Phylodynamic insights into three different phases of the COVID-19 pandemic in Germany

The dynamics of the COVID-19 pandemic have exhibited different patterns over the course of the first two years. In Germany, early outbreaks that facilitated onward spread in the country were followed by seasonal waves of infection, caused by environmental and behavioural changes. After the introduction of different SARS-CoV-2 variants of concern (VOC), these temporal infection profiles were...

20. FAIR Data for Data Science – Contributions from the FUSION Group

Data Science needs data – more precisely it needs FAIR (Findable, Accessible, Interoperable, Reusable)[1], high quality data. Data Engineering approaches for the provision of such data are an important part of data science.

In our work, we develop such approaches mostly in the context of research data. Ongoing activities include the development of BEXIS2, an open source software for...

9. Fine-grained Recognition and Continuous Learning for Biodiversity Monitoring

Dimitri Korsch

Biodiversity monitoring is crucial to understand and preserve our nature. Hence, the computer vision group supports ecologists with AI-based tools for the automated evaluation of images. First, we develop methods for fine-grained species recognition, which is required since the visual appearance of related species complicates traditional recognition methods. We also present methods to enrich...

18. Fluorescence Lifetime Imaging Microscopy (FLIM) Data Analysis by Inverse Modelling

Fluorescence lifetime imaging microscopy (FLIM) provides important information and high-quality images about inter-cellular activity, metabolic state, cellular morphology, etc. [1]. It is a sophisticated imaging approach that relies on the complex curve fitting method by extraction of lifetime parameters. The 'fit-free' deep learning (DL) based lifetime estimation method, which serves as an...

10. Investigating Neural Network Training on a Feature Level using Conditional Independence

There are still unresolved questions regarding the changes in learned representations of deep models during the training process. Gaining a better understanding of this process can assist in validating the training. To achieve this goal, previous research has analyzed training in the mutual information plane. We base our analysis on a method founded on Reichenbach's common cause principle. By...

15. LIBS data analysis

The Laser-Induced Breakdown Spectroscopy (LIBS) technique is widely used to measure the concentration of elements in different types of samples [1, 2]. This study was established to investigate and compare the performance of two approaches, classical regression using Partial Least Square (PLS) and Deep Learning (DL), in predicting the concentration of 24 elements from LIBS spectra. The main...

1. Potential and limits of minimal informative priors for hidden Markov Models to improve parameter inference

Jan Münch (Physiologie 2)

Inferring the complex functional dynamics of ion channels from ensemble currents is a daunting task. We previously addressed this problem by applying a parallelized Bayesian filter to specify kinetic schemes for macroscopic current and fluorescence data leeding to a more accurate likelihood than previous gold-standard algorithms (Muench 2022 eLife 11:e62714).

Using Bayesian statistics...

8. Pre-processing Raman data via deep learning method

Raman spectroscopy is a type of analytical technique that uses the interaction of light with a sample to provide information about its atomic and molecular properties. However, Raman spectra are frequently overshadowed by inconsistencies in baselines and various sources of noise. These defects and contributions to the Raman data must be rectified before identifying or categorizing the samples....

2. Predicting anharmonicity constants using machine learning

Jamoliddin Khanifaev (PhD student)

Considering anharmonic effects is essential to accurately describe vibra-
tional, spectroscopic and thermodynamic properties of molecular systems. Re-
cently, it has been shown that introducing anharmonic formalism into the quantum
cluster equilibrium (QCE) program improves the results of the calculations [1].
It is possible to calculate anharmonic frequencies [2]. However, such...

19. Simplification Models of Scientific and Medical Texts in English and Ukrainian

The language used in scientific and medical texts can be difficult for non-experts to understand. In recent years, there has been a growing interest in developing simplification models to make these texts more accessible to a wider audience. In this presentation, we will explore simplification models of scientific and medical texts in both English and Ukrainian. We will discuss the challenges...

4. Simulation study for artifacts markers in microscopic images

Elena Corbetta (Institute of Physical Chemistry (IPC) and Abbe Center of Photonics (ACP), Friedrich Schiller University Jena, Member of the Leibniz Centre for Photonics in Infection Research (LPI))

Optical microscopy is a powerful and minimally invasive tool for the investigation of biological processes. In this context, processing of images is of utmost importance to improve image quality and sample understanding. However, there is not a standard quantitative approach to evaluate image quality, especially in presence of artifacts. The computation of metrics can provide ambiguous...

22. Test

test

3. The Archive Query Log: Mining Millions of Search Result Pages of Hundreds of Search Engines from 25 Years of Web Archives

Jan Heinrich Reimer (Friedrich-Schiller-Universität Jena)

The Archive Query Log (AQL) is a previously unused, comprehensive query log collected at the Internet Archive over the last 25 years. Its first version includes 356 million queries, 166 million search result pages, and 1.7 billion search results across 550 search providers. Although many query logs have been studied in the literature, the search providers that own them generally do not publish...

6. Understanding the effects of plant diversity on soil temperature stability

Gideon Henrik Stein (FSU Jena, iDiv)

To understand the mechanism that drives the relationship between plant diversity and soil temperature stability, we compare the results of Structural Equation Modeling (SEM) based on domain-expert hypotheses with those of two Causal Inference methods which are entirely agnostic towards possible relationship (The PC algorithm and the CCDr method). While the results are generally consistent, we...

17. Unsupervised Anomaly Detection for Space Gardening

Abstact

The EDEN Roadmap at DLR aims at building a Bio-regenerative Life Support System (BLSS) for future space missions within the current decade.
To ensure the safe and stable operation of the BLSS, the need for automated system monitoring in general and, in particular, robust anomaly detection is apparent.
While the abundance of available methods makes it difficult to choose...

16. Using Advanced Metabolome Annotation Workflow to elucidate Microbial Interactions in Microalgae

Anne-Susann Abel

Microbial communities reveal complex metabolic interactions which critically influence the ecosystem around them. In this instance, we study the interaction in an aquatic ecosystem between the microalgal species Prymnesium parvum, which produces fish killing toxins, and the diatom Skeletonema marinoi, which causes algal blooms. To decipher the roles of individual microorganisms within their...

13. Various Color Transformation Methods

Image registration process involves aligning identical shapes or structures in pairs of related images. In order to accomplish this, one image must be transformed into another. The selection of the appropriate transformation model is essential for accurate registration and it should be selected based on the data. In this study, we focus on color transformation methods and apply two different...