2024

• Discovering causality in time series using feature importance and ensemble models
• Comparing inferential properties and prediction when modelling ordinal data - a simulation study (joined supervision by Moritz Berger (Bonn) and Markus Pauly)
• Simulation and Forecasting Renewable Energy Output: Leveraging Machine Learning Models for Time Series Analysis
• Impact of Missing Data Imputation Methods on Accuracy and Feature Selection of Machine Learning Regression Models (joined supervision by Jakob Schwerter and Markus Pauly within the FAIR project)
• Investigating Fouling in Dairy Ultrafilteration Membranes: A Data-Driven Approach with Piecewise Linear Regression and Machine Learning Models
• Consistency in imputing missing values across various dimensions in official statistics
• Die Prognose von Lieferzeiten in der Transportlogistik mit maschinellen Lernen zur Optimierung der Tourenplanung.
• An Analysis of Causal Inference Methods for Competing Risks (joined supervision by Saraj Friedrich (Augsburg) and Markus Pauly)
• Inference for Competing Risks based on Area between Curves Statistics (joined suupervision by Marc Ditzhaus and Markus Pauly)
• Predicting Station Demand for Washington´s Capital Bike-Sharing System using Spatiotemporal Clustering and Machine Learning
• Active Learning-based Threshold Selection for Online Anomaly Detection in Powertrain Development
• Die Bedeutung von Absatzprognosen für das Operations Management
• Contextual service parts forecasting: An NLP-based approach for First-time fix rate optimization
• A Survival Tree Based on the Copula-Graphic Estimator (joined supervision with Takeshi Emura)

2023

• The Impact of using Text-Data for Prediction Models in Time-to-Event Analysis
• Anomalieerkennung in elektrischen Strömen bei industriellen Pumpen
• Trustworthiness of ChatGPT: A comprehensive analysis of various aspects

• Forecast of Expulsion in Resistance Spot Welding
• Abschätzung von Fertigungspreisen basierend auf geometrischen Ähnlichkeiten mittels maschinellem Lernens
• Topic Detection in Natural Language Processing using Clustering and Semantic Analysis Techniques
• Clustering-based feature analysis in Infineon's power MOSFET portfolio
• Oversampling techniques for epileptic seizure detection
• Statistical Learning-based zero-inflated risk prediction for insurance companies
• Forecasting of continuous and sparse demand of products for inventory management
• Benchmark Data Sets and Simulation Models for Time-to-Event Analyses in Oncology

2022

• AR-Sieve Bootstrap for the Random Forest and a simulation-based comparison with rangerts time series prediction
• Comparison of Modern Machine Learning Techniques for Image Classification
• MultiFANOVA: Multiple Contrast Tests for Functional Data
• Anomaly Detection for Compositional Time Series Data
• Prediction of Weather Forecasting Errors based on Observations using Artificial Neural Networks
• Forecasts for Sales Figures Based on Hierachial Time Series Models
• Warenkorbanalysen mittels Assoziationsregeln und Clusterverfahren
• Vorhersage von Modellfehlern basierend auf Analysefeldern der numerischen Wettervorhersage mittels neuronaler Netze
• Erkennung von Anomalien in hochfrequenten Daten mittels statistischen Lernens

2021

• Prediction Intervals for Time Series: A Comparison of Model-Based and Statistical Learning Techniques (in cooperation with Carsten Jentsch)
• Quantile-based MANOVA: A new tool for inferring multivariate data in factorial designs. (Main supervisor: Marc Ditzhaus)
• Konsistente Vorhersagemodelle für hierarchische Zeitreihen
• Ein Vergleich von uni- und multivariaten maschinellen Lernverfahren am Beispiel der Paketmengenprognose
• Developing an Interpretable Stacking Model for Multilabel Classification.
• Vorhersagen von Aktienmärkten –  ein Vergleich von Zeitreihen- und Künstliche Intelligenz (KI)-Methoden (in cooperation with Carsten Jentsch)
• Machine Learning Based Performance Prediction for Cellular Communication Systems (Co-supervisor: Benjamin Sliwa from the Communication Networks Institute)
• Neuronale Netze in der Baustatistik: Automatsiertes Erkennen von Baustellen anhand von Luftbildern

2020

• A Statistical Analysis of the Klopp-Effect (in cooperation with Andreas Groll)
• Cluster-based permutation tests for neuro-imaging: Pitfalls and a different approach to account for multiplicity
• Einsatz sequenzieller Verfahren im Hyperparameter-Tuning (in cooperation with Andreas Groll; Main supervisor: Daniel Horn).
• Travel Time Prediction with Tree-based Methods
• Short Term Forecasting of Parcel Volumes using Machine Learning and Time Series Analysis

2019

• Prediction of peaks in parcel centers using weather data
• Uncertainty Quantification in Deep Learning

2018

• Deep Learning Techniques and their Application in Finance
• Consistency of Random Forests with Application to Facebook Data
• Theory on k-Nearest Neighbour with Applications to Classification of Bank Marketing Data
• Inference with the Titman Estimator: Confidence Bands, the Bootstrap and Testing Markovianity
• Consistency of random forests with and without sparsity and an application to regional sales forecasting
• Enhancing Cognitive Abilities with Comprehensive Training? A Statistical Analysis with MANOVA and Structural Equation Models

2017

• An Analysis of the Probabilistic Index Model with Application to Neuropsychological Data from Dementia Research
• Modellierung der Lebensdauerstatistik von Lötstellen basierend auf Degradationsdaten
• Statistische Analyse von Längsschnittdaten zu Hochvolt-Batterie-Verschleißgrößen im Feldgeschehen von Fahrzeugflotten
• Eine multivariate Analyse von Kardiologie-Daten: Faktorielle MANOVA und Klassifikation
• Machine Learning Verfahren und ihre Anwendung in der Materialprüfung
• Test for Interaction with Censored Data - Analysis of the GABRIELA-Study and an R Package
• Multiple Imputation in Credit Risk for the Daimler Financial Services AG
• Classification of Cognitive Diseases with Discriminant, Principal Component and Random Projection Analyses

2016

• An Empirical Process based Analysis of the Mean Restricted Residual Lifetime - Theory and Statistical Application
• Untersuchung der Blindbedienbarkeit berührungssensitiver Eingabefelder durch haptische Augmentierung
• Statistische Analyse von Unfällen mit schweren Nutzfahrzeugen in Deutschland
• Entropie bei körperlicher Aktivität: Eine faktorielle Analyse von Wiederholungsmessungen
• Bootstrap-basierte Konfidenzintervalle für Schwangerschaftsausgangswahrscheinlichkeiten unter starker Linkstrunkierung
• Dynamic semiparametric estimation of multistate regression models with application to diabetes survival data
• High-Dimensional Repeated Measures Analysis: Asymptotics and Multiple Samples
• Anwendung der Extremwerttheorie zur Bestimmung des Höchstalters von Steblichkeitskurven

2015

• Diagnostische Gütemaße und ihre Konfidenzintervalle für geclusterte Daten unter Berücksichtigung multipler Gutachter
• Inference for Longitudinal Data Based on Nonparametric Effects

• Maximlian Kertel (2024). Probabilistic Graphical Models in the Manufacturing of Electric Vehicles (external PhD thesis at the BMW Group Center of Excellence for Battery Cells
• Lena Schmid (2024). Statistical Analysis of Tree-Based Ensembles
• Thilo Welz (2022). Robust Covariance Estimation in Mixed-Effects Meta-Regression Models.
• Claus Nowak (2022). Nonparametric correlation-based methods with biomedical applications.
• Lubna Amro (2022). Resampling-Based Inference Methods for  Repeated Measures Data with Missing Values.
• Paavo Sattler (2021). Inference for multivariate and high-dimensional data in heterogeneous designs.
• Burim Ramosaj (2020). Analyzing Consistency and Statistical Inference in Random Forest Models.
• Maria Umlauft (2018). Reliable and Robust Tests and Confidence Intervals for Statistical Functionals with Applications to Biometric and Psychometric Data.
• Sarah Friedrich (2017). Permutation- and resampling-based inference for semi- and nonparametric effects in dependent data.
• Dennis Dobler (2016). Nonparametric inference for multi-state Markovian models with applications to incomplete life science data.