This thesis proposes an attention-enhanced Variational Autoencoder for generating station-specific strong ground motion records. The model encodes three-component PEER NGA-West2 seismic acceleration waveforms as six-channel STFT spectrograms and learns compact latent representations through a convolutional encoder with an attention-based bottleneck. A station-aware latent sampling strategy produces site-specific synthetic recordings from limited per-station data. A structured evaluation framework is introduced in the scope of the thesis. This framework combines time-domain metrics and pseudo-spectral acceleration analysis through intensity-shape binning. Generated records are benchmarked with the evaluation framework against original recordings and SCEC Broadband Platform simulations across Southern California stations.

Date: 30.04.2026 / 14:30 Place: A-212

This thesis presents a systematic evaluation of Static Application Security Testing (SAST) tools. Related studies mostly use synthetic codebases and per-vulnerability evaluation methods. In this study, both synthetic benchmarks and real-world intentionally vulnerable applications are tested against tools, along with per-issue evaluation. Conducted experiments measure various metrics and explain these results with reasons behind them. In addition to quantitative results, qualitative features and internal mechanisms of tools are examined to further explain results and observed performance differences. The results demonstrate the difference between evaluation models and tool effectivenes. Overall, thesis offers practical insights for SAST tool research and selection.

Date: 14.04.2026 / 11:00 Place: Cisco Lab

Somatic mutations drive the transformation of normal cells into cancer. However,distinguishing driver mutations, which confer a selective growth advantage, from the vast background of neutral passenger mutations remains a critical challenge. To address this, we introduce a novel graph-based framework that constructs mutation-centric networks by leveraging longrange genomic interaction data. Our method models genomic intervals as nodes and their long-range interactions or overlaps as edges. Starting from a ’seed’ mutation, the graph expands iteratively, finding overlaps and interacting intervals to capture both local and distal genomic context. This architecture allows us to quantify a mutation’s topological influence, identify complex structural patterns (such as graph cycles), and assess proximity to known driver genes across variable ranges. Furthermore, this approach naturally generates embeddings for individual mutations, enabling the clustering of samples based on mutation profile similarity. Ultimately, by providing a comprehensive, interaction-aware view of the genomic landscape, our framework facilitates more accurate driver identification and improved patient stratification for personalized treatment.

Date: 13.04.2026 / 14:00 Place: B-116

This thesis presents an empirical investigation into the impact of Large Language Model (LLM) assistance on the human factors of software practitioners. Through a controlled within-subject experiment conducted with 30 practitioners, the study compares conditions with and without LLM support using a software requirements specification task. The research utilizes validated psychometric scales and semi-structured interviews to analyze four key dimensions: perceived task complexity, motivation, sense of achievement, and creative self-efficacy. The study provides structured empirical evidence on how LLM tools shape cognitive and psychological processes beyond technical productivity.

Date: 03.04.2026 / 14:30 Place: A-212

This thesis aims to improve demand forecast accuracy through the implementation of machine learning/deep learning applications. SARIMA, LSTM, CNN, and Prophet models are implemented to forecast demand. The models are trained and validated using forward-chaining cross validation method. It applies time series forecasting on a seasonal, unbalanced and intermittent dataset. The effects of selected exogenous indicators and aggregation-disaggregation approaches are examined for further improvement.

Date: 20.01.2026 / 14:30 Place: A-212

This thesis presents an integrated framework for AI-assisted BIM analytics and digital twin development to support indoor navigation and emergency evacuation. Revit-based BIM data are transformed into a unified pipeline by separating geometric representations from structured spatial information and integrated into a Unity-based digital twin. Real-time indoor pathfinding is enabled using the A* algorithm while considering physical evacuation constraints. A large language model (LLM) provides natural-language navigation instructions and supports context-aware user interaction. The system enables interactive visualization and testing and is evaluated through user studies, demonstrating a scalable and human-centred approach for BIM-based digital twin applications.

Date: 14.01.2026 / 14:15 Place: II-06

The problem of distinguishing legitimate domains from DGAs is an ongoing challenge. CTAs use DGAs to bypass traditional detection mechanisms like blacklisting or static signature-based solutions. Since blacklisting all DGA-generated domains is not possible, the automatic detection of DGAs remains a critical necessity. Despite years of research, modern malware continues to evolve, and this makes detection increasingly difficult. This thesis employes a hybrid framework that integrates CNN based n-gram learning and LSTM for automatic feature extraction and Logistic Regression (LR) for final classification. The model input is further enriched by Top-Level Domain (TLD) information in order to improve DGA detection.

Date: 16.01.2026 / 11:00 Place: Cisco Lab

This thesis presents a comparative empirical evaluation of AI-generated front-end code quality. Three contemporary AI coding tools—Cursor, Gemini CLI, and Windsurf—are assessed through controlled experiments involving React and TypeScript applications of varying complexity. The study evaluates both single-prompt and incremental development approaches using industry-standard metrics, including Lighthouse, SonarQube, and ESLint. The findings demonstrate that while AI tools can generate functionally correct applications, code quality outcomes vary significantly depending on tool choice, development methodology, and application complexity. The results provide evidence-based guidance for selecting AI tools and workflows in modern front-end development.

Date: 20.01.2026 / 13:00 Place: A-212

This thesis introduces the Adaptive Reasoning Curriculum (ARC), a standardized framework that uses Large Language Models (LLM) for Automated Curriculum Learning (ACL). Designed to improve generalization in complex Open-Ended Learning (OEL) environments like Neural MMO 2, ARC facilitates efficient skill acquisition rather than environment memorization. Experiments demonstrate that ARC significantly outperforms expert-curricula and existing ACL methods across average return and sample efficiency. By providing open-source adapters and auxiliary tools, ARC establishes a reproducible, accessible standard for integrating LLMs into multi-agent RL, fostering broader innovation in open-ended research.

Date: 14.01.2026 / 13:00 Place: A-212

This study proposes a framework for investigating the alignment between institutional reviewer guidelines and peer review practice. A dataset was constructed from the ICLR 2024 venue on OpenReview. Large Language Model (LLM) pipelines were utilized to extract reviewer inquiries, perform topic modeling, and apply soft classification. These outputs were compared against the ICLR 2024 reviewer guideline to quantify the guideline-practice gap. Generated topics are also mapped to the guidelines from the top ten AI venues. The proposed framework provides a data-driven approach to identify implicit evaluation norms for venue organizers and gives authors and reviewers a data-driven roadmap of implicit evaluation norms.

Date: 14.01.2026 / 14:30 Place: A-212